Peter C. Aichelburg and Roman U. Sexl (Eds.

Albert Einstein
His Influence on Physics, Philosophy and Politics

With Contributions by
Peter G. Bergmann, Hiroshi Ezawa, Walther Gerlach,
Banesh Hoffmann, Gerald Holton, BernulfKanitschneider,
Arthur I. Miller, Andre Mercier, Roger Penrose,
Nathan Rosen, Dennis W. Sciama,Joseph Weber,
Carl-Friedrich von Weizsacker,John A. Wheeler and
Wolfgang Yourgrau

Published under the auspices of the "International Society

on General Relativity and Gravitation"

Friedr. Vieweg & Sohn Braunschweig/Wiesbaden

The papers of W. Gerlach and C. F. v. Weizsacker were translated by
Dr. and Mrs. M. Skopec.

The paper of B. Kanitscheider was translated by R. U. Sexl.

The contribution of R. Penrose is a completely revised version of an

article from "Cosmology Now", British, Broadcasting Corporation.

The copyright of the articles by J. A. Wheeler rests with the author.

Quotations from Einstein by kind permission of the Estate of
Albert Einstein, Otto Nathan, Trustee, New York, N.Y., U.S.A.

All rights reserved

© Friedr. Vieweg & Sohn Verlagsgesellschaft mbH, Braunschweig, 1979
Softcover reprint of the hardcover 1st edition 1979
No part of the material protected by this copyright notice may be reproduced
or utilized in any form or by any means, electronic or mechanical, including
photocopying, recording, or by any informational storage and retrieval
system, without written permission from the copyright owner.

Set by Vieweg, Braunschweig

ISBN 978-3-528-08425-7 ISBN 978-3-322-91080-6 (eBook)

DOI 10.1007/978-3-322-91080-6
 v

Introduction

Dart nun, bei den Heiden, bei diesen wirkiich vorbild-

haften Menschen erscheint uns das Interesse fiir die
Person, fiir den Namen, fiir Gesicht und Gebiirde er-
iaubt und natiiriich.
H. Hesse, "Das Giasperienspiel"

In 1979 the world celebrates the centenary of Albert Einstein's birth.

This offers an occasion to review his life and his scientific work in retrospect,
to survey his importance for our time, and to look forward to future years
of scientific research.
Undoubtedly, Einstein was one of the key-figures in the intellectual
history of our century. He influenced physics and philosophy, as well as
politics. The creation of general relativity is one of the greatest scientific
achievements of our time, as well as the apex of Einsteins's scientific work.
Its full implications for the other fields of physics have become clear only in
recent years. The technological possibilities offered by space research have
enabled mankind to survey the universe for the first time unhindered by the
earth's atmosphere. This has led to new discoveries and has shown that even
some of the far-reaching conclusions derived from Einstein's theory are borne
out by observation. General relativity, which has for a long time been viewed
as an outsider among physical theories because of its mathematical difficulty
and complexity, is considered now to be the prototype of theories in the
fields of elementary particle physics and even solid state physics.
The contributions to this volume attempt to demonstrate Einstein's
influence on the intellectual history of our century. Three of the contributors
have been co-authors of Einstein, several others have had an intense exchange
of ideas with him. The papers collected here demonstrate the far-reaching
influence of Einstein's work and his personality on the physics, philosophy,
theory of science and on the politics of our time.
The first papers survey Einstein's scientific work in the field of relativity
and some of the problems which are of interest in this respect today. The
President of the International Committee for General Relativity and Gravita-
tion, Peter Bergmann, one of Einstein's collaborators, opens this series of
papers with a short introduction to relativity and its confirmation by ex-
periment. Of special interest are Bergmann's remarks on the various attempts
which have been made to generalize relativity and the personal recollections
of the au thor concerning his joint work with Einstein.
VI Introduction

General relativity has led to completely new answers to the old problem
of the structure of the universe. The progress of relativistic cosmology, which
started with Einstein's "Kosmologischen Betrachtungen zur Allgemeinen
Relativitatstheorie", was hindered for several decades by the lack of ap-
propriate observational material. The discovery of the Hubble law in the
twenties led to the idea of an expanding universe which had its origin several
billion years ago in a big bang. It was only the discovery of the cosmic back-
ground radiation by Penzias and Wilson in 1965, however, which gave a
further and probably decisive hint to the reality of the hot and dense initial
phase of the universe. Dennis Sciama shows in his short introduction to
relativistic cosmology how the experimental data slowly led to a more
quantitative and exact history of the cosmic revolution.
It was already in 1920 that Einstein concluded from the field equations
of general relativiry that gravitational waves should exist, which propagate
with the velocity of light. At that time it seemed out of the question to
detect these waves experimentally. All known mechanisms for the generation
and discovery of gravitational waves led to effects which were too small to
be measured. Therefore the scientific community was very sceptical when
Joseph Weber began in 1950 first with theoretical considerations and later
with practical attempts to improve gravitational wave antennas. The decisive
improvement of the sensitiviry of these antennas, which is due to him, has
encouraged other scientists. Currently there are about twenty groups working
on the development and improvement of gravitational wave antennas. Weber
surveys the present state of the art in his paper and shows that quantitative
results may be expected within the next years.
An important source of gravitational waves is the gravitational collapse
of stars. Depending on the mass of a star this collapse can lead either to a
white dwarf, a neutron star or a black hole. While white dwarfs have been
known to astronomy for a long time, the discovery of neutron stars came
only in 1967. A research group at the University of Cambridge found periodic
radio signals at that time, which were emitted by stars. Theoretical con-
siderations showed that neutron stars were the only possible source of this
radiation. Thus the second of the three possible final states of the gravita-
tional collapse of a star had been found. The open and outstanding problem
was now to discover a third possible form, black holes. At first it was not
certain whether these singularities of space-time would actually be formed in
the collapse of realistic stars. Theoretical considerations by Stephen Hawking
and Roger Penrose showed however, that the complete annihilation of matter
in gravitational collapse can be expected not only in improbable idealized
cases, but as a general feature of the gravitational collapse of massive objects.
This result initiated the development of methods for the search for black
holes. The paper by Roger Penrose shows that several astronomical objects
are known today which are likely to contain black holes. It is unfortunate
that Einstein could not live to see this confirmation of the most daring con-
Introduction VII

elusion to be drawn from general relativity. In an imaginary dialogue John

Wheeler tries to reconstruct Einstein's possible reaction to this discovery.
Einstein's contributions to physics were not restricted to the field of
relativity, but dealt also with quantum theory and thermodynamics. The
creation of the light quantum hypothesis in 1905 was one of the decisive steps
towards quantum theory. It initiated the wave-particle dualism which turned
out to be one of the most difficult problems in interpreting the quantum
theory. Einstein himself and some of the other leading physicists of the time
remained sceptical towards the quantum theory. In a famous paper "Can the
Quantum-mechanical Description of Reality be Considered Complete" by
Einstein, Podolski and Rosen, he laid down the reasons for his dissatisfaction
with the orthodox Copenhagen interpretation of quantum mechanics. This
paper led to many comments and Nathan Rosen discusses his present position
in his contribution.
Einstein's contributions to quantum theory are closely connected with
his research on thermodynamics, which is analyzed by Hiroshi Ezawa.
This analysis illustrates how Einstein's scientific work was guided by the
fundamental postulate of the simplicity and unity of the universe. These
postulates are taken up from a different point of view by Arthur I. Miller in
his contribution analyzing the origin of special relativity and in Gerald
Holton's studies of Einstein's approach to theory formation. It is interesting
to see the similarities in the different treatments of one subject by various
authors: the physicist Ezawa is mainly interested in Einstein's influence on
the third development of thermodynamics and shows how Einstein tried
again and again to "understand" Planck's law, thereby working his way
towards quantum mechanics. Miller's study of the origin of relativity is
written in a completely different style. The methodology of the history of
science (which might seem to the physicist to be overly accurate in some
places) proves that Einstein's research was opposed to the scientific fashions
of his time and shows the variety of problems which have been solved by
special relativity. On the other hand, Holton's paper leads to an analysis of
Einstein's work from the point of view of the philosophy of science. Holton
discusses Einstein's method of theory construction on the basis of a letter
from Einstein to his friend Solovine. This letter reveals the full importance
of the jump which leads from the field of experience to the formulation of
the axioms of a theory: "There is no logical path to these laws; only intuition,
resting on sympathetic understanding of experience, can reach them."
The philosophy of science on which Einstein's papers are based is also
analyzed in the contribution by Bernulf Kanitscheider. It becomes apparent
how much Einstein differed in his scientific work, but also in his methodologi-
cal reflections, from the positivistic attitudes of the philosophy of science of
his time. In an intuitive way Einstein's ideas anticipated several decades of
developments in the philosophy of science, which led to a more liberal
attitude towards the introduction of theoretical entities.
VIII Introduction

Carl-Friedrich von Weizsiicker's contribution deals with the connection

between physics and philosophy, which is brought out by Einstein's work,
but also with his importance for politics. "Einstein was a physicist not a
philosopher. But the naive directness of his questions was philosophical".
This naivite is also characteristic for Einstein's attitude towards politics, in
which he was increasingly involved in his later years. The paper by Banesh
Hoffmann shows how Einstein was won over to the idea of Zionism only
slowly after 1920, but was deeply concerned with Jewish affairs thereafter.
The concern for the fate of the Jews and the fear of a victory of Nazi-
Germany in the Second World War were also decisive for Einstein's famous
letter to President Roosevelt, in which he pointed out the possibility of an
atomic bomb. This letter, written in August 1939, and a second one in
March 1940, were crucial for the funding of the first controlled chain reaction
and the atomic bomb.
The influence wh"ich political developments have even in the fields of
pure and most abstract science such as general relativity and gravitation, is
borne out by Andre Mercier's account of the history of the GRG-organization.
The International Committee for General Relativity and Gravitation, which
tried to co-ordinate the international research efforts in the field of general
relativity, repeatedly had severe problems in overcoming the barriers be-
tween different political systems and in encouraging a genuine world-wide
collaboration.
The last contributions to this volume deal with personal recollec-
tions. For Walter Gerlach Einstein's contribution to quantum theory was
most important. In the early history of quantum mechanics one of the
central problems was the real existence of photons. In a variety of experi-
mental arrangements one tried to clarify whether the emission and absorption
of photons was an instantaneous process or whether it corresponded rather
to the continuous emission of waves, which was postulated by classical
theory. The arguments were so heated, that frauds were used in order to
prove the opposing points of view.
John Wheeler's recollections deal mainly with relativity. A visit at
Einstein's house in 1953 and Einstein's last lecture, which took place on
April 14, 1954, illustrate Einstein's search for a unified field theory. The
detailed notes taken during Einstein's last lecture give an insight into the
style of his lectures, in which he tried to present a synthesis of many fields
of physics. Finally, Wolfang Yourgrau describes some personal meetings
with Einstein which illustrate the lighter side of academic life.

***
Introduction IX

We are especially indebted to the authors whose contributions have

made this publication possible. Furthermore, we are obliged to Dr. O. Nathan
who gave us permission to quote from the Einstein Estate. We thank
Mrs. J. Aichelburg, as well as Dip!. Ing. E. Oberaigner, and Doz. Dr. A. Wehr!
for their valuable help in revising the manuscripts, and Miss E. Klug and
Mrs. F. Wagner for their untiring typing efforts. We especially want to
mention the pleasant cooperation offered by the Vieweg Publishers, who
were always willing to accommodate our wishes, despite the pressing dead-
lines.

Peter C. Aichelburg Roman U. Sexl

Vienna, November 1978

x
Contributing Authors

Peter G. Bergmann
Professor of Physics at Syracuse University, USA; born in Berlin, he emigrated to the
USA and became in 1936 assistant to Einstein at the Institute of Advanced Study in
Princeton; numerous articles on special and general relativity; first attempts to quantize
the gravitational field; author of the book "The Riddle of Gravitation"; President of the
International Society on General Relativity and Gravitation.

Hiroshi Ezawa
Professor of Physics at Gakusuin University, Japan; research on quantum field theory
and quantum statistics; author of the book "Who has seen the Atom?"

Walther Gerlach
Professor (Emeritus) for Physics, University of Munich, Bundesrepublik Deutschland;
essential contributions to experimental quantum theory (Stern-Gerlach experiment for
directional quantization); publications on radiation, spectroscopy, magnetism and history
of science.

Banesh Hoffmann
Professor for Mathematics, Queens College, New York, USA; co-worker of Einstein and
member of the Institute for Advanced Study; publications and research on relativity in
particular on the motion of matter in a gravitational field, quantum theory, applications
of tensor analysis to electrical engineering; author of a biography of Albert Einstein.

Gerald Holton
Professor of Physics and History of Science at the Massachusetts Institute of Technology,
USA; research in history of science and philosophy of science; author of "Thematic origins
of Scientific Thought: Kepler to Einstein" and "The Scientific Imagination: Case Studies".
Contributing Authors XI

Bernulf Kanitscheider
Professor for Philosophy of Science at the University of GieBen, Bundesrepublik Deutsch-
land; studies on the concept of geometry and its meaning for physics; author of "Geo-
chronometrie und Wirklichkeit" and "Vom absoluten Raum zur dynamischen Geo-
metrie".

Arthur I. Miller
Assoc. Professor for Physics at Lovell University, USA; interdisciplinary research in the
history of 19th century science.

Andre Mercier
Professor (Emeritus) for Theoretical Physics, also Philosophy at the University of Bern,
Switzerland; research and numerous publications on mathematical methods of theo-
retical physics, theories on the origin of the earth, the concept of time, theory of knowl-
edge; author of the book "Analytical and Canonical Formalism in Physics" and others;
former Secretary-General of the International Society on General Relativity and Gravita-
tion.

Roger Penrose
F.R.S., Rouse Ball Professor for Mathematics, University of Oxford, England; formulated,
together with S. Hawking the first theorems on the existence of space-time singularities
in general relativity; research on black hole physics, techniques of differential topology
in relativity.

Nathan Rosen
Professor of Physics at the Israel Institute of Technology, Israel; colleague of Einstein's
at the Institute for Advanced Study in Princeton; research in relativity and unified field
theory, quantum theory, thermal diffusion, fundamental particle theory, gravitation and
cosmology.

Dennis W. Sciama
Fellow of All Souls College, Oxford, England; research in astrophysics, cosmology and
general relativity; author of several books e.g. "Modern Cosmology".

Joseph Weber
Professor of Physics at the University of Maryland, USA; he developed the first detectors
for gravitational waves; publications and research on relativity, microwave spectroscopy,
irreversibility; au thor of "General Relativity and Gravitational Waves".
XII Contributing Authors

Carl-Friedrich von Weizsa'cker

Professor and Director of the Max-Planck-Institut zur Erforschung der Lebensbedingungen
der wissenschaftlich-technischen Welt, Starnberg, Bundesrepublik Deutschland; numerous
publications and books on physics, philosophy and peace research; his book "Zum Welt-
bild der Physik" was translated into ten languages, several distinctions, Max-Planck Medal
and Pour Ie Merite.

John A. Wheeler
Professor in Physics at the University of Texas, Austin, USA; he worked many years in
Princeton and was in friendly contact with Einstein; during the Second World War Advicer
on atomic energy projects, co-worker on the Manhattan Project; creator of "Geometro-
dynamics" where matter is described as topological properties of space-time; decorated
with the Einstein Medal and several other awards; author of the books "Geometro-
dynamics", "Einstein's Vision", "Gravitation" and others.

Wolfgang Yourgrau
Professor of Physics at the University of Denver, USA; studied in Berlin with Schrodinger,
Einstein and v. Laue; research on quantum theory and the theory of measurement;
editor of "Foundations of Physics".
 XIII

Contents

Peter G. Bergmann

The Development of the Theory of Relativity 1

Dennis W. Sciama

Cosmology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 17

Joseph Weber

Gravitational Radiation 25

Roger Penrose

Black Holes .............................................. 33

John A. Wheeler

The Black Hole: An Imaginary Conversation with Albert Einstein . . . .. 51

Nathan Rosen

Can Quantum-Mechanical Description of Physical Realty

Be Considered Complete .................................... 57
XIV Contents

Hiroshi Ezawa

Einstein's Contribution to Statistical Mechanics. . . . . . . . . . . . . . . . . .. 69

Arthur I. Miller

"On the History of the Special Relativity Theory" 89

Gerald Holton

Einstein's Model for Constructing a Scientific Theory .............. 109

Bernulf Kanitscheider

Einstein's Treatment of Theoretical Concepts .................... 137

Carl Friedrich v. Weizsiicker

Einstein's Importance to Physics, Philosophy and Politics . .......... 159

Banesh Hoffmann

Einstein and Zionism ....................................... 169

Andre Mercier

Birth and Role of the GRG-Drganization and the Cultivation

of International Relations among Scientists in the Field ............ 177
Contents xv

Walther Gerlach

Reminiscences of Albert Einstein from 1908 to 1930 .............. 189

John A. Wheeler

Mercer Street and other Memories ............................. 201

Wolfgang Yourgrau

Einstein - and the Vanity of Academia . . . . . . . . . . . . . . . . . . . . . . . .. 231

 1

The Development of the Theory of Relativity

Peter G. Bergmann

As early as World War One Einstein himself wrote a popular introduc-

tion to the theory of relativity which was then published by Vieweg. Even
today the study of this work, which in my opinion can hardly be surpassed,
can be recommended to the interested layman. Certainly, a lot was added in
the course of the next sixty years which might even affect the basis of the
theory. This arcticle has necessarily to be rather short but it is based on the
assumption that the reader will consult Einstein's as well as other writings
on the subject.

From Ether to the Theory of Relativity

The classical conceptions of space and time were sha?ed by Galilei and
Newton and represent an essential modification of the ideas of the Middle
Ages. The classical concept of space is that of the three-dimensional Eucli-
dean manifold to which the concepts of the straight line and the plane are
basic. I t is assumed that all points in Euclidean space are in principle equi-
valent, and this assumption includes the transition from the geocentric
model of the Middle Ages to the model of the universe in which earth, our
solar system, and even our galaxy are by no means a privileged center of the
universe but only objects of which one is convinced that there exist numerous
of a similar kind which are dispersed everywhere.
Likewise, time represents a one-dimensional manifold wherein no point
(= instant) stand out among the others. Specifically, an instant of the crea-
tion of the universe is ruled out.
A system of reference is a combination of space and time, in which the
concepts of rest and motion of material objects are well defined. A special
kind of systems of reference are inertial systems to which the principle of
inertia applies: a body isolated from all interactions with other physical
objects stays in its condition of rest or uniform motion. It is a physical
hypothesis to postulate such inertial systems, for all assumptions concern-
ing physical space and physical time are not postulates of pure mathematics
or geometry - they allow many other models - but those of physics, or
more generally, of natural science.
2 Peter G. Bergmann

If we designate an interval of time with the symbol T, the vector point-

ing from one event to the next one with 5, and the velocity of the second
system of reference with respect to the first one with;' then the quanteties
T and 5' of the second system are in relation to the first inertial system:
T' T
--*, -;:t--*
5 = .) - v T. (1)

This "classical" assumption is called the Galilean transformation between

two inertial systems.
It can be easily shown that the laws of Newtonian mechanics, if valid
in one inertial system, apply to all other inertial systems under the assump-
tion (1). Furthermore, the relations form a group of transformations: they
transform from a second to a third system of inertia in such a way that their
relative velocities tr, 1b, ... add as vectors. Thus the Newtonian laws and the
Galilean transformations (1) represent a logically self-contained system.
The motive for changing this system did not come from within but
from the outside due to the growing knowledge about the electromagnetic
field, the laws of which are not consistent with those of classical mechanics.
As a consequence of Maxwell's field laws the electromagnetic waves
propagate in a vacuum with the speed of ligth (c = 300,000 km/s) in all
directions. Supposing there existed one inertial system to which this strictly
applies, then it would follow from (1) that in another inertial system the
velocity of transmission of electromagnetic waves would depend on the
direction and would deviate from the universal value c.
It was tempting to postulate that among all inertial systems one was
distinguished due to the fact that the velocity of light is independent of
direction. Since the earth is strictly speaking not an inertial system (due to
its revolution around the sun), one had to assume it would be possible to
measure the directional dependence of the velocity of electromagnetic waves
if sufficiently refined methods of observation were found.
The famous Michelson-Morley experiment tried to determine this
effect. It was, as the disappointed Michelson later stated, a failure. Neither
this experiment nor the following experiments were successful in detecting
differences of the speed of light in different directions. Modern versions of
the Michelson-Morley experiment even show that the speed of light is the
same in all directions with an accuracy of 3 cm/s.
Naturally many theoreticians' tried to do justice to this situation.
Lorentz and Poincare assumed that the measured speed of light was indepen-
dent of direction in all inertial systems. Fitzgerald demonstrated that this

P. Ehrenfest gave a survey of the ~ituation at that time (1912) in his inaugural address
in Leyden: "Zur Krise der Lichtiither-Hypothese" ("On the Crisis of the Light-ether
Hypotheses").
The Development of the Theory of Relativity 3

Gestatten Sie, in einigen grellen Strichen das BUd

zu skizzieren, das sieh so ergibt: Der Atherwind st6rt
den Ablaul der Prozesse, mit denen der Experimentator
operiert; derselbe Atherwind verdirbt aber auch -
wenn wir UllS so ausdrucken diirfen -=- die MeBinstru·
mente des Experimentators: er delormiert die MaB-
stabe, verandert den Gang der Uhren und die Feder...
kraft in den Federwagen usw. Fur alles da~ sorgen

Zur Krise der jene Grundhypothesen, insbesondere auch die Hypo-

these, daB die Bewegung durch den Ather die An-

Lichtather -Hypothese ziehungskriifte zwischen den Molekiilen verandert. Und

wenn nun der Experimentator die durch den Atherwind
gf'st6rten Prozesse mit seinen Instrumenten beobachtet.
die derselbe Atherwind verdorben hat, dann sieht er
Rede exakt das, was der ruhende Beobachter an den unge-
gehalten beim Antritt des Lehramts stOrten Prozessen mit den unverdorbenen Instru-
an der Reichs-Universitat zu Leiden menten beobachtet.

von
Die Grundhypothesen der '904-Arbeit sorgen daltir,
daB auch bei allen anderen Atherwinrlexperimenten
Prof. Dr. P. Ehrenfest immer wieder die Wirkung des Atherwindes vor dem
Experimentator verborgen bleibt.
Sie sehen: die '904-Arbeit von Loren tz zeigt
tinen ma,gllchen Auswe-g aus der Krise, in die die
Atlwrhypothese geraten war.
Aber niehl alle Physiker glaubten sich mit dieser
Lasung der Krist' zufrieden geben zu kannen.
Wir komm('n damit an die beiden Standpunkte
heran, die Einstein im Jahre 1905 und Ritz im
] ahre 1908 publizierten. Leider mussen wir uns ver..
sagen im Rahmen dieser Rede, eine Besprechung
dieser Standpunkte zu versurhen. Wir begnugen uns,
j ene Ztige in ihnen hervorzuheben, die ihre Stellung
Berlin
innerhalb der Atherkrise markieren.
Verlag von Julius Springer
Das negative Ergebnis aller Atherwind-Experi-
19I3
mente £Uhrt beide Autore!l zur Uberzpugung, daB es
uberhaupt keinen Ather gibt. Der Raum
zwischen den K6rpern sei leer. Die Elektronen der
Karper werfen einander durch diesen leeren Raum hin~
durch die eleklromagnetischen Impulse und das Licht
zu. Kurz. beide Autoren betonen. daB im Gegensatz
zur Athertheorie von Lorentz ihre Theorien wieder
an die Emissionstheorie von Newton anknupfen.

Fig, 1
Ether theory and relativity were compared in an elegant paper by P. Ehrenfest in 1912.
4 Peter G. Bergmann

would be the case if in systems of reference moving against the universe with
the speed v, all scales and all fixed objects were shortened at the rate of
VI - V2 /C 2 without changing their dimensions vertically to the directions of
motion. In other words, an inertial system of "absolute rest" should exist,
but all other inertial systems could not be distinguished from this privileged
system due to deformation of the measuring instruments which were also in
motion. There would be an apparent equivalence of all inertial systems not
corresponding, however, to the geometric-dynamic reality.
The young Albert Einstein was dissatisfied with this situation. In ac-
cordance with most other theoreticians he assumed that by means of obser-
vation no inertial system could be distinguished from all others and assumed
in addition (such as Poincare, whose works in this field were probably un-
known to Einstein) that the equivalence was of a basic nature and searched
for the physical consequences. In 1905 he published his paper "Zur Elektro-
dynamik bewegter Karper" ("On the Electrodynamics of Moving Bodies") in
which he postulated for the first time that two events taking place simultane-
ously in one inertial system, but at different places, neednot happen simul-
taneously in another inertial system. The reason for the "relativity of simul-
taneity" is that simultaneity can only be defined through the exchange of
signals and no signals can travel with a speed higher than c.
Thus it follows from the relativity of simultaneity that time intervals
depend on the inertial system in which they are measured and that spatial
relations are also modified. The new formula for two events connected by a
vector parallel to, the relative motion of two inertial systems is the Lorentz
transformation (as named by Einstein):

(2)
S'= 1 (S-vT).
jl-v 2 /c 2
As regards light signals, if the two events are the emission and the reception
of a light signal, the relation S = cT or S = - cT (depending whether the
signal moves in the direction v or in the opposite direction) implies the identi-
cal relation between S' and T'. Thus the speed of light is the same in both
systems of inertia and is independent of the direction of transmission.
Obviously, equations (1) and (2) cannot be "correct" despite the fact
that both systems are without logical contradiction within themselves. The
properties of "real" time and "real" space must be determined by means of
observations and experiments. All experiments which were conducted in the
second half of the 19th century in order to determine these questions ap-
proached the border of the experimentally feasible. As, for instance, the speed
of the earth around the sun is 30 km/sec, the ratio ofthis speed to the speed
The Development of the Theory of Relativity 5

of light is 1: 10,000 or 10- 4 . Since the square of these relations enters into
the Michelson-Morley experiment, the minimum accuracy of the experiment
must bei 10-8 or one part in a hundred million. This is one reason why parti-
cularly this experiment was repeated again and again and with ever more
refined methods of measurement. (See diagram on Ether-drift Experiments.)
Today the special theory of relativity based on equations (2) can be
tested by completely different methods. All particle accelerators which are
used today are based in their fundamental designs on the laws of special
relativity. They could not function if the old equations (1) represented
physical reality. The reason is that for the accelerated particles, electrons or
protons, the ratio vic is not extremely small as is the case with the speed of
earth, but is close to one, i. e. these particles nearly reach the speed of light.
One may conclude that the special theory of relativity gives rise to
a substantial revision of the concepts of space and time, which fuses space
and time even more intimately than classical physics did. However, special
relativity employs a number of principles of former conceptions. The most
essential is that among all imaginable systems of reference there is still one
privileged class, viz. inertial systems. to which the first of the three New-
tonian laws applies: in the absence of external forces a material body remains
unaccelerated. In Minkowski's space-time continuum the inertial systems
playa similar role as the cartesian co-ordinates do in Euclidean geometry.

Kennedy Thorndike
X Michelson X
@Miller
xMichelson
Morley XMorley
XJoos XMU-Grup
Miller

til

"E
c 10
XCedarholm et.a!.

XCialdea

X Is sak et.a!.

1880 1900 1920 1940 1960 1970 Year

Fig. 2
The upper limit for the velocity of the earth with respect to the ether was measured
repeatedly between 1880 and 1970. Only the measurement by Miller, indicated by ® ,
lead to a positive result for the ether drift. This result was explained later by temperature
gradients in the laboratory (Diagramm by R. Mansouri)
6 Peter G. Bergmann

The Path to the General Theory of Relativity

The general theory of relativity goes beyond the special one substanti-
ally. Whereas special relativity had its origin in the problems of electro-
magnetism and the transmission of electromagnetic waves, general relativity
is concerned with the field of gravitation. It is fundamental to gravitation
that the force exerted on a material body is proportional to its mass. In other
words, all material objects undergo the same acceleration under the influence
of a gravitational field. No other real field (of force) has this quality with the
exception of the so-called fields of inertia.
Fields of inertia or forces of inertia are accelerations with respect to
non-inertial systems of reference. Centrifugal forces or Coriolis forces are
typical forces of inertia. The normal explanation for the independence of
accelerations from the properties of affected bodies is that these are kine-
matic and not dynamic effects which disappear at the transition into an
inertial system.
Certainly there exist no systems of reference in which true gravitational
forces disapperar everywhere, i. e., in an extended space-time area. As far as
this is concerned fields of gravitation are different from fields of inertia.
There are, however, systems of reference, in which gravitational forces
disappear locally. In this respect Jules Verne was wrong when he had his
space travelers sense gravitation inside a projectile although their shell was on
a ballistic orbit. (The only force-free point he accepted between earth and
moon was one at which the gravitational forces of the two world bodies
cancelled.) Verne should have foreseen that his travelers around the world
were unable to sense gravitation because their shell was accelerated in the
same way as they themselves. In our time every astronaut and cosmonaut
attests this.
However, if one tries to extend such a free-falling system of reference,
one will not be successful in regions, the range of which is about the same as
the distance from the earth. One can observe the existence of a gravitational
field in such regions by means of tidal effects even if floating in space and
surrounded by impenetrable clouds so that one cannot practise optical
astronomy. Even then one cannot measure the gravitational field-strength
(local effects are not sufficient to determine an inertial system but only a
free falling system) but only its gradient, i.e., its spacetime change.
A number of alternative formulations of these properties of gravitation
exist. One formulation distinguishes between the inertial mass of a body,
relating the applied force and the resulting acceleration of a body on the one
hand, and its gravitational mass on the other, i.e. the strength by which it
attracts other bodies or is attracted. This formulation states that these two
kinds of mass are universally the same for all bodies. The equality of gravita-
tional and inertial mass has been examined again and again due to its univer-
sal importance and with an ever increasing degree of exactness. The most
The Development of the Theory of Relativity 7

recent determination was given by R. Dicke and confirms the validity of this
law with an accuracy of 10- 12 . It ranks among the most exact quantitative
confirmations of a law of nature carried out up to now.
Another formulation states that although the existence of a field of
gravitation can be determined through observation of the field gradient, it
is impossible to separate the relative contributions of the gravitational field
and the possible field of inertia. This formulation comes to the conclusion
that uniquely defined inertial systems (in which the field of intertia di-
sappears) exist only in the absence of gravitational fields. It is more difficult
to convert this definition into an exact experimental procedure than the first
one; however, it comes closer to the conceptual root of general relativity.
Both formulations describe a property of the field of gravity which is con-
sidered correct at present. It is called the principle of equivalence which was
well known to Newton but was not fused with the formulation of mechanics.
Newtonian physics prefers inertial systems as the privileged systems of
reference par excellence.

The End of Euclidean Geometry

The mathematical formulation of general relativity is based on the

modern version of differential geometry which was created by Gauss and
Riemann. These mathematicians discovered in the 19th century that a mani-
fold can deviate substantially in its internal structure from Euclidean geo-
metry by having an internal curvature. This is frequently explained by using
the example of the surface of the sphere. Imagine two-dimensional organisms
which live on the surface of smooth sphere and decide to transport a vector
parallel to itself along a closed curve. "Vector" denotes in this two-dimen-
sional geometry an object which has no components perpendicular to the
surface of the sphere. Let the surface of the sphere be described by longitudes
and latitudes as usual in geography. Starting at the equator with a vector
which is directed exactly towards east we displace it along the equator by e
degrees. Then the vector still points towards the east at the end. But now let
us displace the vector along a meridian to the north pole. There the vector
will be perpendicular to the meridian. In returning to the starting point along
a different meridian we notice that the two meridians enclose .an angle () at
the pole. Therefore the vector will form an angle (90 0 + e) with the new
meridian. Upon returning to the starting point the vector will form an
angle e with its original direction, although it has been transported parallel
all the time. A general theorem posed by Gauss states that the angle between
the two vectors is always equal to the solid angles spanned by the surface
around which the vector has been transported. (In our case this surface is a
spherical triangle.) This angle is frequently called the "spherical excess" since
it is also equal to the angle by which the sum of the angles in a spherical
triangle exceeds 180 0 .
8 Peter G. Bergmann

We owe the generalization of differential geometry of arbitrarily curved

surfaces to Gauss. Riemann extended Gauss's concepts to manifolds of arbi-
trary dimensions. It turned out that Gauss's concept of curvature remains
applicable provided one admits closed paths around surfaces of arbitrary
orientation and does not restrict the parallel vector to lie within the surface
considered. As a consequence of this, twenty curvature components can be
defined in every point of a four-dimensional manifold.
Einstein's main idea was that, mathematically speaking, a gravitational
field has to be equivalent to a curvature of the space-time continuum and
that one had to learn to conceive the property of the gravitational field as
properties of the Gauss-Riemann curvature. All other physical phenomena
take place in this curved space-time.
The detailed formulation of this idea took Einstein many years of hard
work. His fundamental paper on special relativity is dated 1905 and the
fundamental papers on general relativity were published in 1915 and 1916.
In general relativity there exists no inertial system, and an arbitrary system
of co-ordinates, may be used for the description of nature. The laws of gravi-
tation restrict the curvature components in such a way that only ten of the
twenty components remain unspecified. The other ten components are fixed
by the sources of the gravitational field; i.e. mainly mass density of gravitat-
ing matter.
In every point of a curved two-dimensional surface one can construct a
tangential plane with Cartesian co-ordinates. In the same way one can con-
struct a plane tangential space-time in every space-point at a given time of a
curved space-time. This tangential space-time is the freely falling system of
reference which corresponds in one space-time point only to the physical
space-time. Thus it cannot be extended to arbitrarily large space-time regions.
In limited areas, however, real space-time and tangential space-time are
almost identical and one can study physics there in the "usual" (that is,
specially relativistic) way.
On the whole this is the physical and geometrical content of general
relativity. While the measurable properties of space and time pre-existed in
special relativity and were nothing but the eternal stage for the events of
nature, the geometrical properties of space and time are themselves parti-
cipants of the dynamic course of events in general relativity, and are in-
fluenced by everything that happens and themselves influence all events. In
many respects the break between general relativity and earlier theories of
space and time is much deeper than the break between special relativity and
classical physics. Geometry and dynamics are united to form a whole which
was named "geometrodynamics" by Wheeler.
The Development of the Theory of Relativity 9

The Success of Einstein's Theory

Already in his first paper Einstein indicated some possibilities for test-
ing general relativity experimentally. A consequence of the theory is that a
light ray is deflected from its straight course in a strong gravitational field.
This prediction was confirmed experimentally shortly after the First World
War. The results of the expedition conducted by Eddington made Einstein
a famous public figure although the accuracy of the experimental measure-
ment was only about 30 %. This accuracy was not improved on during the
following decades.
It was only around 1960 that improved methods of radar technology,
satellites, and atomic clocks opened up a new era of experimental accuracy
in the confirmations of general relativity.

j'oJectory of photon

Fig. 3
The geometry of space in the vincinity of the sun is represented here by a warped surface,
embedded into an Euclidean space. A light ray passing close to the sun is deflected and
delayed by the curvature of space.
10 Peter G. Bergmann

Red shift and time dilatation. A light ray rising in a gravitational field
loses energy and its frequency decreases. At first one tried to measure this
red shift by investigating the spectral lines of the sun and of white dwarfs.
But it was only the Mossbauer-Effect which brought the breakthrough in
1965: Pound and Snider were able to measure the red shift of spectral lines
in the earth's gravitational field. A relative frequency shift AlJ/lJ = 10- 15 was
confirmed with an experimental accuracy of roughly 1 %.
The red shift of spectral lines is closely connected with another pre-
diction of general relativity. It states that clocks are slow in the vicinity of
heavy masses such as the earth. The development of atomic clocks with an
accuracy of 10- 14 made a measurement of this effect feasible (Table 1).
Light deflection. The development of large scale radio-interferometers
led to improved measurement of light deflection. The observation of radio
waves emi tted by quasars or satellites has increased the limit of accuracy to
about 1 %.
Perihelion advance. One of the most important predictions of general
relativity is the advance of the point of a planet's orbit closest to the sun
(perihel). It was known even before general relativity that the planet mercury
shows a perihelion advance which cannot be explained by pertubations of its
orbit due to other planets.
Improved measurements of this effect were possible during recent
years. The exact positions of various planets were determined by reflecting
radar beams from planetary surfaces. With this technique the perihelion
advance of mercury was determined with an accuracy of better than 1 % and
agreement between theory and experiment was obtained.
The Shapiro experiment. In 1965 a new test of general relativity was
suggested by I. I. Shapiro. A radar beam emitted from the earth is reflected
by Venus. The echo is received on earth and the round-trip travel time is
measured. According to general relativity there is a time delay larger than
that predicted by Newtonian theory. There are two reasons for this time

Table I Accuracy of measurements of relativistic time dilatation

Author Type of experiment Accuracy

Brault (1962) D 1 line of solar spectrum 5%

Pound & Snider (1965) Mossbauer effect C57 1%
Jenkins (1966) GEOS-1 satellite 10%
Hafele & Keating (1972) Cesium clocks in airplane 14 %
Alley (1975) Cesium and Rubidium clocks 1%
in airplane
Vessot (1975) Maser in scout rocket 0.2 %
Bailey et al (1977) Lifetime of muons 0.1 %
The Development of the Theory of Relativity 11

2,95

2,75 -
( 29)

2,55
u Eddington Freund-

-0
0
U
2,35
Dyson
Davidson
lich /
Kluber
(51) Weighted Mean
!l>
2,15 \ Braun
III of Data
f(7) \
-
c

.c
1,95 --------f~8J __
III
..Cl
1,75
(71) " Bei Sbroeck
E -Campbell
.....J
Trumpler 'Einstein'S prediction
1,55
1
(11)
1,35

1919 1922 1929 1936 1947 1952 1959

Year

Fig. 4
The deflection of light was measured between 1919 and 1952 with an accuracy of about
20 % (Diagramms 4-7 adapted from a paper by J. P. Richard).

30279

Fig. 5
Earth The arrangement of radio antennas used for the
+ measurement of light deflection.
12 Peter G. Bergmann

1,2 Experiment = 0.975 :!:O,024 -

Theory

1,1 t- -

r
I,D
t-- - - - c-- - - ,r - - -=± - - - -f-- --

0,9

0,8
I-

t- Seielstad
---v--'
I
Sramek Sramek Shapiro et al Sramek,
-

Hill Fomalon
0,7 I- -

1969 1970 1971 1972 1973

Fig. 6
The results of radio-interferometer measurements of light deflection.

Experiment
Theory

1,1 o

1.0 - --i - - --f~---f--­ o

---0--

0,9
f
Shapiro ,M6 ~7, M9
i
v
et al Others Anderson Sh~piro Anderson
(1974) (69-74) (1973) (1974) (1974) (1978)
Light Bending Time Deloy

Fig 7
Measurements of the delay of radio waves (Shapiro experiment). M and V denote Mariner
and Viking experiments resp ..
The Development of the Theory of Relativity 13

delay. Firstly, clocks are slow in the vicinity of the sun, so that the velocity
of light is decreased there. Furthermore, the path of the radar beam follows
the warped space in the vicinity of the sun and is delayed thereby.
Improvements of radar technology have made possible measurements of
this effect with an accuracy of about 0.15 %. Thus all classical tests of relati-
vity are confirmed today - after decades of uncertainty - with an accuracy
of about 1 %.

Alternatives to General Relativity

Unified field theory. General relativity leads to a geometrical language

for gravitational fields. All other force fields which are known to physics
remain in their original form, which has apparently little resemblance with
geometry. Not only Einstein but many of his contemporaries considered
this duality to be esthetically unappealing and searched for new ways which
would make all physical fields manifestations of geometry. These programs
were called unified field theories. The most important programs, historically
speaking, can be characterized as follows:
a) Conformal geometries. In special as well as well as in general relativity
there exists an invariant measure (strictly speaking this measure is defined in
general relativity only in the tangential manifold). This requirement can be
weakened by considering all measures which differ only by a scalar factor to
be equivalent. This weakening or generalization is called conformal geometry.
In it curvature has to be defined slightly differently than in Riemannian spaces
and one obtains richer structures. Particularly Hermann Weyl tried intensely
to formulate new physical theories based upon conformal geometries which
were to include the gravitational field as well as the electromagnetic one. No
other fundamental fields were known at that time (1918). Weyl never
succeeded, however, in formulating natural laws in a convincing manner and
his attempt remained unsuccessful.
b) Five-dimensional geometries. In 1921 Kaluza proposed to enrich
geometry by adding another dimension. Thereby the number of curvature
components is increased to fifty. One can attempt to understand the four-
dimensional nature of the observable physical universe by restricting the
five-dimensional geometry in such a way that physical fields are not variable
in the fifth dimension. This type of restriction is similar to the consideration
of static fields in ordinary physics but is to be taken as a general restriction
here.
Kaluza's original proposal has been modified during the course of the
newer history of physics by (among others) Einstein, Bargmann and the
author of this article. They have tried to replace the static fifth dimension
by conditions of periodicity, hoping that quantum phenomena could be
described in this way. This hope has turned out to be erroneous.
14 Peter G. Bergmann

The so-called scalar-tensor theories of gravitations represent another

variation of Kaluza's idea. General relativity considers the gravitational field
to be a tensor, while the electromagnetic field is described by a (four-dimen-
sional) vector, the components of which are the electrostatic and the vector
potential. By relatively small modifications of Kaluza's starting point one
can complete a five-dimensional metric (measure) by adding a scalar field
which has been considered cosmologically significant by the proponents of
this theory. At the present time there are no indications for the physical
reality of this additional field which has in any case no resemblance with any
of the elementary-particle fields known today.
c) Geometry with torsion. In Riemannian geometry the metric deter-
mines the parallel transport of vectors and thus the curvature of the mani-
fold completely. By weakening the link between the metric and the curva-
ture, one obtains a richer geometrical structure. This proposal was made
originally by E. Cartan and was taken up later by Hehl, Trautman et al. One
hopes to understand the "spin" of elementary particles in this way on a
more fundamental level than in conventional theories. No final opinion can
be formulated concerning these theories since they are rather recent.
d) Super-symmetries. In order to understand the families ("super-
multipletts") of elementary particles which have been found experimentally,
one has postulated internal symmetries which should exist in addition to the
space-time symmetries (such as parity). Several researchers have attempted
very recently to merge internal and external (i.e. space-time) symmetries
into a "super-symmetry") and to unify this in turn with the laws of gravita-
tion as formulated by general relativity. The results of these attempts which
started in the middle of the seventies are called "super-gravity". No final
opinion can be formulated concerning these theories.
Every judgment of the merits of unified field theories is necessarily
subjective. In the case of a complete and mature theory one generally agrees
upon the empirical facts it can explain, which facts seem to contradict the
theory, and which are outside the frame of validity. In the case of theories
which are relatively new and speculative, no such statements are possible if
only because the physical interpretation of the formal results of a theory are
often not unique. The basic theme of all unified field theories is firstly, that
the unified theoretical explanation of the embarrassingly rich manifold of
physical events should be possible. Secondly, such a theory cannot be de-
duced (arrived at by induction) from the empirical facts but has to correspond
to the fundamental principle of conceptual simplicity and can therefore be
found only by means of creative thought. It is obvious that the theory arrived
at in this way has to be tested experimentally a posteriori and has to be
proven in this manner.
Even if one feels sympathic towards this attitude one has to admit that
the "creative act" in inventing a new theory has to use clues from the known
and established realm of physics if the new ideas are to be more than pure
fancy.
The Development of the Theory of Relativity 15

It appears to me that since the birth of general relativity new experimen-

tal facts, especially in the field of elementary particle physics, have accumu-
lated at such an unbelievable rate that nobody can tell at present which facts
a unified field theory would have to account for. It appears, however, that
we understand today why all attempts undertaken in the twenties and thirties
in complete ignorance of the huge number of elementary particles known to
us now had to be failures. Only now have we a first glimpse of an empirical
symmetry pattern for newly discovered elementary particles, and a hope
that a new and more successful period of theory creation might start within
our life time. To me personally it seems that Einstein's famous statement,
"God is sublime but not malicious", is justified even today, since, despite
all new and fundamental discoveries nature remains rich enough to prevent
its complete and definitive unrevealing. But this is a personal opinion which
will not be shared by all my colleagues.
Einstein's Philosophical Attitude. During the years from 1936 to 1941
when I was lucky enough to be one of Einstein's young collaborators,
Einstein stated frequently his fundamental attitudes towards physics. His
opinions have also been presented especially in the volume, "Albert Einstein:
Philosopher-Scientist ", which was published in the "Library of Living Philo-
sophers" in 1949 on the occasion of his 70th birthday. Since I do not posses
any notes about Einstein's oral statements but have to rely on my memories
from four decades ago, I can and must not claim any authenticity.
In his youth Einstein was strongly influenced by Mach. When I met
him he did not accept Mach's philosophical attitudes any more. He was
convinced of the objective existence of the universe, an existence which
did not depend on the presence of observers being aware of themselves. To
explore this universe and TO understand it is the task of natural science.
As soon as man comes to terms with the external world on an intellec-
tual basis he does this within the frame of a system of concepts which,
however primitive, he has created himself. With statements such as, "When a
ball is released it always falls to the earth", he supposes that we know what
is meant by a ball, that we accept certain types of possible movements to be
separate classes of the behaviour of material bodies, and finally that we
know what we mean by "earth". The sentence given above relates these
already existing concepts. In maintaining that a ball will "always" fall to
earth we extrapolate a necessarily finite series of attempts and observations.
We understand only very little why our universe is such that these extrapola-
tions are actually successful; in Einstein's words: "The greates miracle is that
there are no miracles."
This regularity of nature, that is, the existence of a continuously
increasing number of causal relations known to us ("if .'. always ... , then ... ")
enables us to explore nature systematically. In this respect the existence of
laws of nature is an unavoidable pre-condition for the existence of science.
According to Einstein the statistical laws of quantum mechanics (to the dis-
16 Peter G. Bergmann

covery of which he has made important contributions) cannot claim to be

ultimate truth but are based on a partial description and knowledge of the
state of a physical system. Up to his death Einstein tried in vain to overcome
the statistical character of quantum mechanics. He was ready to admit the
logical completeness of quantum theory but he doubted its ultimate truth.
The last four decades of Einstein's life were devoted to the search for
a unified field theory. It is likely that he has investigated almost all types
of such theories since the world of his ideas was seemingly inexhaustible.
Nevertheless he was not convinced that the idea of a continuous and diffe-
rentiable field could bring about decisive progress for theoretical physics. He
believed enthusiastically in the strict (non-statistical) causality in nature. But
he was ready to admit the existence of structures other than those of fields·
which could, for instance, be built up from discrete, not continuous elements.
He felt that his experience would enable him to make progress with field
concepts rather than with anything else.
Together with Infeld and Hoffmann Einstein completed general relativity
during the years which I spent at Princeton. Together with his collaborators
he proved that the field equations of this theory determine the orbits of the
sources of the gravitational fields such as the bodies of our solar system.
Today we know that every field theory without preferred systems of refer-
ence has this property, but not specially relativistic theories.
There are a number of other papers out of Einstein's later years which
contain essential contributions to the understanding of general relativity. Up
to his death he combined a sense for the essential with analytical brilliance
and creative force. Among the great scientists of the 20th century Einstein
stands out as the theoretician who has contributed fundamental ideas to a
large number of fields out of a unified view of nature.
 17

Cosmology
Dennis W. Sciama

Introduction

It is arguable that the most important scientific discovery of the 20 th

century is that one can discuss the whole universe in a rational way. After
all, what could be more important than understanding everything that there
is? The key technical tool which permits such a rational discussion is Ein-
stein's general theory of relativity. For the first time this theory provides
us with the means to represent the layout of the universe in space and time
in a manner which is also dynamically selfconsistent. Newtonian theory,
with its emphasis on the fundamental role of world-wide inertial frames of
reference, cannot do this. It is true that in the 1930's so-called Newtonian
models were constructed, bearing a close similarity to the relativistic models,
but this was only possible if it was permitted for inertial frames centred on
different points to be accelerating relative to one another. This is not strictly
part of the Newtonian concept. Moreover, it was not possible to give a satis-
factory discussion of the propagation of light in these Newtonian models,
and since light (and radio waves) is the main tool of the astronomer in ex-
ploring the universe, this defect is catastrophic.
The power of general relativity to represent the whole universe was
first demonstrated by Einstein [1] in 1917, shortly after his field equations
had been adumbrated. At that time the expansion of the universe was un-
.known, and it was natural for Einstein to construct a static model. This he
could do only by modifying his field equations by introducing an extra term
- the cosmological term - to provide an effectively repulsive force to bal-
ance the attractiveness of gravity, so permitting the whole universe to be
static. Later he was to call this step the biggest blunder of his life [2].
He introduced in this paper two further ideas which came to have a
permanent influence on theoretical cosmology. The first was, perhaps ob-
vious, but still of fundamental importance. In order to obtain an exact
solution of his field equations, he constructed a model in which the contents
of the universe were smoothed out into a uniform material medium which
was both homogeneous- having the same properties at all points - and iso-
tropic - having the same properties in all directions. Thus was born the first
highly symmetrical model of the universe. Later we shall see that, on a large
scale, the universe does appear to be highly homogeneous and highly iso-
tropic.
18 Dennis W. Sciama

The second idea was by no means obvious, and this was to exploit the
non-Euclidean geometry of his theory by constructing a model in which
space at anyone time is finite but unbounded - the three-dimensional an-
alogue of the surface of a sphere. The topology of space as well as its geometry
was thus brought into the discussion.
The next major step was taken in 1922 by a Russian meteorologist
Alexander Friedmann [3]. He showed that by not insisting that the universe
be static one could construct a whole family of homogeneous and isotropic
models, with or without the cosmological term. These models would contain
systematic motions of expansion or contraction, and because of the symme-
try assumptions the relative velocity of recession or approach of two regions
would be simply proportional to the distance between the two regions con-
cerned (except at very large distances where extra relativistic effects would
come in e.g. because relative velocities approaching that of light would be
achieved). These non-static models of the universe were established two
years before the astronomers made the decisive observational discovery
that the spiral nebula in Andromeda lies outside the Milky Way system.
This was the first step in the discovery of the observational universe, and this
is the question which we must next discuss.

The Universe in Observation

Many spiral nebulae had been observed already in the last century,
and a controversy developed amongst astronomers as to whether these
nebulae were members .of the Milky Way system or were extremal systems
or Milky Ways in their own right. By 1917 the velocities of several of these
nebulae along the line of sight had been measured by means of the Doppler
effect, and it was known that most of the nebulae were receding from the
sun and at much higher velocities than was usual for individual stars, meas-
ured in hundreds of kilometres per second rather than in tens. In 1924 Ed-
win Hubble [4] took his first decisive step when he showed definitely that
the spiral nebula in Andromeda lies outside the Milky Way. He and other
astronomers then set up a major programme of measuring the distances and
motions of the nebulae, or galaxies as they came to be called. He developed
the picture of the universe more or less as we have it today, with galaxies
occurring in groups and clusters of various sizes, our own Milky Way, for
example, belonging to the Local Group of ab'out 20 galaxies. On a large
scale, if one averages over, say, hundreds of galaxies, there appeared to be
no marked deviations from homogeneity and isotropy.
We begin to have formed here, then a possible picture of the whole
universe in which the ultimate building brick is perhaps a cluster of gal-
axies. But Hubble's most decisive contribution came with his discovery
of the systematic expansion of this system of galaxies. In 1929 he pro-
Cosmology 19

pounded for the first time the famous Hubble [5] law, that the velocity
of recession of a galaxy is proportional to its distance from us. This was
7 years after Friedmann's discovery of the uniform non-static models of
general relativity. The observational astronomers had at last caught up
with the theorists!
In the ensuing years both observation and theory were systematised,
and by the mid-thirties they came into a definite form which can be re-
garded as the ending of an epoch. In 1936 Hubble [6] published his book:
The Realm of the Nebulae, and in the same year H. P. Robertson [7] and
A. G. Walker [8] gave a systematic discussion of all the homogeneous and
isotropic models of general relativity. Those models without a cosmolog-
ical term in the field equations all began with a singular moment of infini-
te density a finite time ago. Some of them then expanded for ever, while
others eventually re-contracted into another singularity of infinite density.
The distinction between these two cases arises from the total amount of
matter and radiation in the universe, through the gravitational effect of
this material. The dividing-line between the two cases is given by a model
which just expands for ever but in which the rate of expansion tends to zero
as time goes on. This intermediate model was particularly recommended by
Einstein himself and by W. de Sitter [9] in a joint paper written in 1932,
and is now always referred to as the Einstein-de Sitter model. To this day,
we do not know whether the universe will in fact expand for ever.

The 3 K Cosmic Black Body Background

The ensuing 30 years or so were mainly years of consolidation. Perhaps

the most interesting idea in this period was the 1948 suggestion of H. Bondi
[10], T. Gold and F. Hoyle [11] that, despite its expansion, the universe
might be in a steady-state. Their idea was that matter might be created contin-
ually in the universe at such a rate that its mean density remained constant.
There would then be evolution in any localised region of the universe, but
not in the universe as a whole. In particular there would be no question
then of a singular origin to the universe a finite time ago. According to
Hubble's estimates at that time, the singular moment of the conventional
theories occurred unconfortably recently, indeed more recently than the
best current estimates of the age of the Earth, the Sun and the Milky Way.
Hubble's estimates have since been revised upwards, and there is now no
time-scale difficulty (although only just!) Nevertheless the steady-state
theory would have remained of fundamental significance despite its contra-
dicting general relativity had it not been for the accidental discovery in
1965 of excess radio noise reaching the Earth from outside at a wavelength
of 3 centimetres.
20 Dennis W. Sciama

(km/s)

10 4

Fig. 1
The Hubble law: The recession
velocity of Galaxies is proportional
to their distances from the earth.
(The conversion of apparent lu-
minosities into distances is based on
the value H = 1.6 . 10- 18 s -1 for the
Hubble constant.)
10 25
v(km)

This apparently innocuous discovery by A. A. Penzias and R. W. Wil-

son [12] has turned out to be the most important observation made about
the universe as a whole since the discovery of the Hubble Law. As later ob-
servations showed, when one measures the intensity of this radiation at dif-
ferent wavelengths one finds that the spectrum is that of radiation which has
reached thermal equilibrium at a definite temperature (namely 3 OK) and
thus conforms to Planck's law for black body radiation. The implications
of this discovery for cosmology are manifold, and it is sad that Einstein
never knew about it, particularly because some of his best early work in
physics concerned the constitution and fundamental properties of black
body radiation.
The possible cosmological significance of Penzias and Wilson's dis-
covery was immediately pointed out by R. H. Dicke, P. J. E. Peebles, P. G.
Roll and D. T. Wilkinson [13] even before its spectrum was established
observationally. They took the conventional view that the universe has
evolved ,to its present state from an initially dense configuration. The new
point is that this dense configuration would also have been very hot, since
as the universe expands any black body radiation field would cool down.
The 3 K of today then represents 109 K at a time of one hundred seconds
after the "big bang" origin of the universe. In fact G. Gamow [14], R. A.
Alpher and R. C. Herman [15] had already suggested earlier that the big
bang might have been very hot, and that this heat might have survived
until today. Their argument had been based on a study of the nuclear
reactions which might have occurred in the hot early phase, building up
Cosmology 21

Fig. 2
Penzias and Wilson in front of
the microwave antenna used for
their discovery of the cosmic
black body radiation. In 1978
they were awarded with the
Nobel prize for physics.

heavier elements out of lighter ones, and so perhaps accounting for the
distribution of elements we see today in the Milky Way. By 1965 this ar-
gument had been forgotten, mainly owing to the success of the alternative
theory in which the heavier elements are built out of the light ones in
supernova explosions.
This situation was rather ironic since the motivation behind this super-
nova picture was the steady-state theory with its abolition of the hot dense
phase of the universe. This theory has itself been made extremely implau-
sible by the discovery of the 3 K background. The point is that in the uni-
verse as it exists today any excess radiation could not be thermalised on the
time scale available. We need a denser universe for such thermalisation
to be possible. If we calculate the last possible moment during the evolution
of the universe at which the radiation could have been thermalised we find
that it is only 300 years after the big bang, when the universe was lOIS
times denser than it is today. Thus when we measure the spectrum of the
background to be thermal we are in effect observing directly the result of
processes which occurred in a universe at least 101~ times denser than the
present universe. Clearly the steady-state theory is ruled out by this argu-
ment.
22 Dennis W. Sciama

A second irony is that the supernova theory remains the most convinc-
ing explanation for the origin of the heavy elements, but not for the lightest,
namely helium and deuterium. It is now fairly certain that most of the
helium we observe today was made by nuclear reactions about 100 seconds
after the hot big bang when the temperature was 109 K. It is less certain,
but fairly likely, that most of the present deuterium was also made then.
This would be of particular interest because of observed abundance of
deuterium can, in the simplest form of the theory, be made in the early
universe only if its mean density is so low that the universe would be destined
to expand forever. However, this conclusion can be avoided by suitably com-
plicating the theory.
The final feature of the 3 K background which we want to mention is
its isotropy. Observations have shown that the temperature does not vary
with the direction of observation [16] (except for a small effect attributed
to the motion of the Earth through the background) by as much as 1 part in
3.000. This is by far the most accurate measurement ever made in cosmology,
and it tells us many things. Here we would mention only two. The first
is that on a large scale the universe must be highly homogeneous and iso-
tropic. Otherwise there would be gravitational effects on the background as
it propagates to us which would show up in an anisotropy of the tempera-
ture distribution. Thus the assumption of the homogeneity and isotropy
of the universe, which were originally introduced mainly to enable exact
solutions of Einstein's field equations to be obtained, now turn out to be
good representations of reality. Why the universe should be so symmetrical
remains a mystery.
The second consequence of the isotropy of the background is that we
can use it in conjunction with the Hawking-Penrose [17] singularity theorems
to show that according to classical (that is, non-quantum) general relativity,
the big bang origin of the universe really was singular. This means that
actually the theory itself has broken down at this early stage. We thus
reach a crisis for general relativity, but one determined not externally,
but by the theory itself. How this crisis will be resolved is unknown. The
least radical solution would be that in transcribing general relativity to make
it conform with quantum theory the singularity would be eliminated.
However, more radical measures may have to be taken. Since we do not yet
know how to quantise general relativity the matter remains unsettled.
Clearly the early universe presents us with a physical laboratory so extreme
that even the greatest theory of space time and gravitation so far devised
is not not able to cope with it. I am sure that Einstein in this centenary
year would have wanted this fact emphasised rather than the triumphs of
his great theory.
Cosmology 23

References

[1] A. Einstein, "Kosmologische Betrachtungen zur allgemeinen Relativitatstheorie",

Preuss. Akad. Wiss. Berlin. Sitzungsber. p. 142, 1917.
[2] A. Einstein, as quoted by G. Gamow "My World Line", p. 44 (Vilzing Press, New
York) 1970.
[3] A. Friedmann, "Uber die Kriimmung des Raumes", Z. Phys. 10, p. 377, (1922).
[4] E. P. Hubble, "NGC 6822 a remote stellar system", Astrophys. J. 62, p. 409, (1925).
[5] E. P. Hubble, "A relation between distance and radial velocity among extragalactic
nebulae", Proc. Nat. Acad. Sci. U.S. 15, p. 169, (1929).
[6] E. P. Hubble, 'The Realm of the Nebulae", New Haven, Yale Univ. Press, 1936.
(7] H. P. Robertson, "Kinematics and world structure", Astrophys. J. 82, p. 248,
(1935).
[8] A. G. Walker, "On Riemannian spaces with spherical symmetry about a line and
isotropy in general relativity", J. Math. Oxford Ser. 6, 81 (1935).
[9] A. Einstein, W. de Sitter, "On the relation between the expansion of the uni-
verse", Proc. Nat. Acad. Sci. 18,213 (1932).
[10] H. Bondi und T. Gold, "The steady-state theory of the expanding universe", Mon.
Not. R. Astron. Soc. 108, p. 252 (1948).
[11] F. Hoyle, "A new model for the expanding universe", Mon. Not. R. Astron. Soc.
108, p. 372, (1948)
[12] A. A. Penzias und R. W. Wilson, "A measurement of excess antenna temperature at
4080 Mc!s", Astrophys. J 142, p. 419 (1965).
[13] R. H. Dicke, P.]. Peebles, P. G. Roll und D. T. Wilkinson, "Cosmic-black-body ra-
diation", Astrophys. J. 142, p. 414 (1965).
[14] G. Gamow. "The evolution of the universe", Nature 162, p. 680, (1948).
[15] R. A. Alpher und R. C. Herman, "Evolution of the universe", Nature 162, p. 774
(1948).
[16] G. M. Smoot, M. V Gorenstein und R. A. Miiller, "Detection of anisotropy in the
cosmic blackbody radiation", Phys. Rev. Lett. 39, 14, p. 898 (1977).
[17] S. W. Hawking und R. Penrose, "The singularity of gravitational collapse and cos-
mology", Proc. R. Soc. London A 314, p. 529 (1969).
 25

Gravitational Radiation~:-
] oseph Weber

Introduction

Einstein's general theory of relativity predicts that changes in the

gravitational field are propagated with the speed of light, and that certain
kinds of sources may radiate away energy and momentum as gravitational
waves.
Observation of such radiation is important for a number of reasons.
There are several modern theories of gravitation, and these predict different
detailed structure of the waves. Gravitational radiation therefore provides
an imp ortant tool for testing different theories of gravitation. The absorp-
tion of gravitational radiation by matter is very much smaller than the
absorption of other forms of radiation. Therefore, observations of gravita-
tional radiation may provide astronomical observations of phenomena which
either do not emit significant amounts of light and radio or X-radiation, or
cannot be observed because emitted light is obscured by opaque matter.
Gravitational waves provide an entirely new channel of information about
the universe.
The most direct and perhaps most satisfactory way to observe gravita-
tional radiation would be to follow the pioneering work of Hertz, Roentgen,
and Barkla in producing laboratory sources of such radiation and apparatus
for detection. The great weakness of the gravitational force made such exper-
iments beyond the reach and grasp of the conceivable technology of the
two decades 1950-1970. We are readily aware of the effects of gravitation
when masses as large as the earth, sun, and moon are involved. However in
the laboratory we have available electrons to generate light, and somewhat
similar processes would generate gravitational waves. However the gravita-
tional force which one electron exerts on another at rest is Hr43 smaller
than the electrical force. This ratio of forces carries over to the radiation
rates. Gravitational radiation by electrons is also a factor 10-43 weaker
than electromagnetic radiation.

* Research supported in part by NASA Grant II NSG 7196-S 1 and in part by National
Science Foundation Grant PHY 77-14818
26 Joseph Weber

Modern elementary particle physics provides a description of gravita-

tion as transmitted by gravitons which have the same relationship to gravita-
tional waves that photons have for electromagnetic waves. Here again the
weakness of the interaction has thus far made it impossible to carry out
gravitational quantum physics experiments corresponding to those in other
branches of elementary particle physics.
For these reasons we decided to rely on the large masses contained in
astronomical sources, and devoted most of our efforts to exploration and
development of suitable antennas to receive radiation from such sources.

The Gravitational Radiation Antenna

General relativity theory unifies geometry and gravitation physics.

The geometrical description of gravitation is that of the curvature of four
dimensional spacetime. If we have a region free of gravitational fields the
geometry is Euclidean. For a triangle constructed of light rays (Fig. 1) the
sum of the angles is 180 degrees. For a triangle surrounding a body such as
the sun (Fig. 2) the sides of the triangle are curved, and the sum exceeds 180
degrees. Such a space is "curved". A gravitational wave propagates this kind

Figure 1
Triangle made of light rays in Euclidean (gravitation free) Space.

Figure 2
Triangle made of light rays in space curved by the gravitational
field of the sun.
Gravitational Radiation 27

of curvature. For gravitational waves having plane surfaces of constant phase

normal to the propagation direction, the constant phase surfaces are Eucli-
dean two dimensional spaces. If we look in the direction of propagation of
the waves at a given instant we would see a distribution of space time curva-
ture as indicated in Figure 3. Regions of positive curvature are separated by
half a gravitational wave length from regions of negative curvature. There-
fore, one way of searching for gravitational radiation is to measure directly
the sum of the angles of small triangles made of light rays. The possible
sources known in 1960 were the closest double stars with periods of a few
hours. For these sources the difference between the sum of the angles of a
small triangle and 180 degrees, is 10-40 radians. Such an incredibly small
quantity was well beyond the technological limitations of that period.
Clearly the gravitational radiation antenna would have to employ some
other phenomenon.

A U A U
0;
u
0
0..
'"
E
0
VI
a

{j A (j A
1:-
Q)
E
0

'"
01
'0
'"

A (j A U
01
c
0
.c
u

~
0
c.
'" E
E '"
i-=~

Direction of propagation
(change of geometry at different places but same time)

Figure 3
Regions of space alternating from convex to concave in path
of gravitational wave.
28 Joseph Weber

Such a phenomenon is the tidal effect of gravitation. The gravitational

field of the moon varies from place to place on earth. As a result there are
differential forces which distort the shape of the earth and give rise to
tides in both the solid earth and the oceans. In 1958 I was able to prove
by detailed analysis using Einstein's equations that similar effects are pro-
duced on an extended body by a gravitational wave. Measurement of the
tidal effect does in fact measure the spacetime curvature. Einstein's equa-
tions permit an exact solution to be obtained for the gravitational radiation
antenna. The tidal effect of gravitational waves produces strains in a solid
which vary with time in accordance with the frequency of the waves. Meas-
urement of these strains measures the space time curvature. For certain
kinds of elastic bodies the strain produced is enormously greater than the
quantities involved in the triangle method discussed earlier.
The analysis also implied that the distance between free masses would
be affected by gravitational radiation. This suggested use of a Michelson
interferometer with the mirror distance changes to detect the gravitational
radiation. Such a device was first developed by our Maryland student,
Dr. Robert L. Forward.

Cylinder Antennas for the Kilohertz Region

In 1959 we decided to construct antennas in the form of large metallic

cylinders (Fig. 4). If a gravitational wave is incident from the direction
shown, the length of the cylinder will undulate and the resulting strains are
converted to electromagnetic signals by means of piezoelectric crystals. The
techniques of modern electronics are then employed to observe these strains.
Gravitational waves are only one of many effects which can produce
strain and output signals from a gravitational radiation antenna. Ground vi-
brations, and many kinds of noise will give spurious output signals. In
order to isolate the antenna from ground vibrations, the cylinder is suspend-
ed at a node and the suspension rests on an array of acoustic filters. Electri-
cal shields are employed to attenuate local electromagnetic disturbances.
However, even if all external disturbances are sufficiently reduced, there
remains an internal noise background due to the heat induced motion.
For a one ton mass at room temperature, this corresponds to a relative
end face displacement amplitude of about 10- 14 centimeters. Thus for an
antenna more than one meter long, we may observe changes in length
much smaller than the diameter of an atomic nucleus, if all disturbances
other than internal ones can be sufficiently attenuated by isolation.
The successfully instrumented antenna of Figure 4 operated un-
attended at remote locations for long periods.
Gravitational Radiation 29

Figure 4
University of Maryland kilohertz frequency gravitational radiation antenna.
30 Joseph Weber

VOLUME 22, NUMBEa 24 PHYSICAL REVIEW LETTERS 16JUNII969

EVIDENCE FOR DISCOVERY OF GRAVITATIONAL RADIATION·

J. Weber
Department of Physics and Astronomy, University of Maryland, College Park, Maryland 20742
(Received 29 April 1969)

Coincidences have been observed on gravitational-radiation detectors over a base llne

of aheut 1000 km at Argonne National Laboratory and at the University of Maryland.
The probability that all of these coincidences were accidental is incredibly small. Ex-
periments Imply that electromagnetic and seismic effects can be ruled out with a high
level of confidence. These data are consistent with the conclusion that the detectors are
being excited by gravitational radiation.

Some years ago an antenna for gravitational ra- array Is a new set of windows for studying the
diation was proposed. 1 This consists of an elas- universe.
tic body which may become deformed by the dy- Search for gravitational radiation In the vicinity
namic derivatives of the gravitational potentials, of 1660 Hz. - A frequency In the vicinity of 1660
and its normal modes excited. Such an antenna Hz was selected because the dimensions are con-
measures, precisely, the Fourier transform of venient for a modest effort and because this fre-
certain components of the Riemann curvature ten- quency is swept through during emission in a
sor, averaged over Its volume. The theory has supernova collapse. It was expected that once
been developed rigorously, starting with Ein- the technology was refined, detectors could be
stein's field equations to deduce' equations of designed for search for radiation from sources
motion. Neither the linear approximation nor the with radio or optical emiSSion, such as the pul-
energy-flux relations are needed to describe sars. A knowledge of the expected frequency and
these experiments, but their use enables discus- Q of a source enormously increases the probabil-
sion in terms of more familiar quantities. All Ity of successful search.
aspects of the antenna response and slgnal-to- However, occasional signals were seen at 1660
noise ratio can be written In terms of the curva- Hz and small numbers of coincidences were ob-
ture tensor. The theory was verified experiment- served on detectors" • separated by a few kilo-
ally by developin~ a hl~h-frequency source' and meters. To explore these phenomena further,
producing and detecting dynamic gravitational larger detectors were developed. One of these II
fields In the laboratory. now operating at Argonne National Laboratory.
Several programs of research are being car- My definition of a coincidence is that the recti-
ried out. One employs laboratory masses In the fied outputs of two or more detectors cross a
frequency range 1-2 kHz.' Another Is concerned given threshold in the positive direction within a
with expected gravitational radiation from the specified time interval. For the present experi-
pulsars. 5 Some designs for such antennas sug- ments the time Interval was 0.44 sec. The mag-
gest a pulsar detection range approaching 1000 nitudes of the outputs at a coincident crossing en-
pc. A third class of antennas employs the quad- able computation of the probability that the coin-
rupole modes of the earth, 1 the moon, and plan- cidence was accidental. Observation of a number
ets' for the range 1 cycle/h to 1 cycle/min. This of coincidences with low probability of occurring

J._~~~-L~~~~ '.~~~~~~~-L

COINCIDENCE TIME MARK J -ARGONNE DETECTOR ___ __:.... _-,-_' __ -L~- .' L I' //

COINCIDENCE TIME MARK J MARYLAND DETECTOR

FIG. 2. Argonne National I.aboratory and University

of Maryland detector colnciden,e.

Figure 5
Gravitational Radiation 31

Antennas of Improved Sensitivity

The internal noise background of the antenna can be reduced by going

to very low temperatures. Greater sensitivity may also be obtained through
use of exotic materials such as large single crystals of sapphire and silicon.
Research programs to develop such antennas are being carried out by a num-
ber of laboratories throughout the world.

Search for Gravitational Radiation at Kilohertz Frequencies

Search has been carried out using the antennas already described.
Since the expected signals are rare, and large local disturbances may occa-
sionally occur, it is necessary to search for coincidences of output for an-
tennas at two locations. Such investigations have been carried out. The
Maryland group. observed significant numbers of signals. A number of
other research groups found nothing. A group in Tokyo and a group in
Munich and Rome found smaller numbers of signals. These observations
are very controversial.
With improved sensitivity, it is hoped that there will be successful
detection of radiation from sources as far away as the Virgo cluster of
galaxies, by observatories throughout the world.

Search for Radiation at Low Frequencies

For very low frequencies, the earth and moon may be employed as
elastic solid antennas. The earth may have oscillations excited at a fre-
quency of one cycle every fifty-four minutes, and higher overtone frequen-
cies. The moon may have oscillations excited at a frequency of one cycle
every twenty minutes, and higher overtones. Einstein's general relativity
theory predicts that only certain overtones will be excited by gravitational
radiation. The Maryland group observed the surface acceleration of the
earth for a number of years, and also observed the surface acceleration of
the moon making use of equipment emplaced by the Apollo 17 astronauts.
These observations have thus far produced no evidence that either earth
or moon is excited by gravitational waves.
 33

Black Holes
Roger Penrose

About 6,000 light years away, in the constellation of Cygnus the Swan,
lies the blue supergiant star HDE 226868. Its mass exceeds that of our
sun by a factor of about thirty, and its radius by a factor of nearly twenty-
five. This in itself is nothing especially unusual. Many other stars of a simi-
lar nature are known. But once every five and one-half days, HDE 226868
circles in orbit about an invisible companion. It is this mysterious compan-
ion which concerns us here - with a mass one-half that of HDE 226868
but utterly tiny, its radius apparently being only about fifty kilometres.
The companion of HDE 226868 is now believed by many astronomers to be
a black hole - a bizarre consequence of the physical laws embodied in
Einstein's theory of general relativity. The identification of the companion
of HDE 226868 as a black hole is not yet quite certain, but looks highly
probable at the present time.
There are other objects in the heavens which some astronomers claim
are likely also to be black holes and it may be that a final definitive judge-
ment on the existence of black holes could lie, instead, with one of these.
It is thought that black holes might lie at the centres of galaxies or of globu-
lar clusters of stars. There is quite good evidence for the existence of a vast
black hole perhaps 100,000,000 km in diameter, at the centre of M 87, a
galaxy ejecting a huge jet of luminous gas.
But however it comes, such a discovery would be aT: event of the utmost
importance to present-day physical theory. For theory predicts that black
holes should exist and should occur sometimes as the end-point of stellar
evolution. If black holes were not eventually found to exist then this would
point to some drastic revision necessary in the theory. On the other hand
their existence will also pose fundamental problems for theory and I shall
attempt to elucidate some of these later in this chapter.

What is a black hole?

For astronomical purposes it behaves as a small, highly condensed

dark 'body'. But it is not really a material body in the ordinary sense. It
possesses no ponderable surface. A black hole is a region of empty space
(albeit a strangely distorted one) which acts as a centre of gravitational
34 Roger Penrose

attraction. At one time a material body was there. But the body collapsed
inwards under its own gravitational pull. The more the body concentrated
itself towards the centre the stronger became its gravitational field and the
less was the body able to stop itself from yet further collapse. At a certain
stage a point of no return was reached, and the body passed within its
'absolute event horizon'. I shall say more of this later, but for our present
purposes, it is the absolute event horizon which acts as the boundary surface
of the black hole. This surface is not material. It is merely a demarcation
line drawn in space separating an interior from an exterior region. The inte-
rior region - into which the body has fallen - is defined by the fact that no
matter, light, or signal of any kind can escape from it, while the exterior
region is where it is still possible for signals or material particles to escape
to the outside world. The matter while collapsed to form the black hole
has fallen deep inside to attain incredible densities, apparently even to be
crushed out of existence by reaching what is known as a 'space-time singu-
larity' - a place where physical laws, as presently understood, must cease
to apply.
One very puzzling question concerning black holes is the following:
how can a black hole exert gravitational attraction on other bodies when
its content is completely shielded from the outside world by the absolute
event horizon? How is it possible for the gravitational field of a collapsed
body to escape, when no information or signals can? The solution to the
problem is as follows: the gravitational field does not "escape" but corre-
sponds with that of the body before the collapse. When the body in the
centre is crushed in the course of a collapse, the outer gravitational field
cannot simply be cut off. In such a case a signal coming from the black
hole would be necessary in order to "give notice" to the gravitational field,
as to when the body disappears. The outer field indicates in no way at all
what is taking place in the interior. After the collapse of the body it is
more accurate to imagine the black hole as a stable configuration of the
gravitational field itself. The field makes no further use of the body that
produced it!
Since the black hole acts as a centre of attraction it can draw new
material towards it - which once inside can never escape. The material
thus swallowed contributes to the effective mass of the black hole. And
as its mass increases the black hole grows in size, its linear dimensions
being proportional to its mass. Its attractive power likewise increases, so
the alarming picture presents itself of an ever-increasing celestial vacuum
cleaner- a maelstrom in space which sweeps up all in its path. But things
are not quite so bad as this. We are saved by the very minuteness of black
holes - a fact which results from the smallness of the gravitational con-
stant.
Black Holes 35

HOF 226 68(sup.rglant stor)

Rotation

,
.
Cygnus X... (block hoi.)
, ....
' ....... I'

Rotation
Sun, drawn to
'am. Kal.- 0
Fig. 1
The gravitational field of the black hole distorts the supergiant star out
spherical shape and drags material from it.

To see this , let us return to our picture of HDE 226868 (Fig. 1). Ac-
cepting the most recent figures for the dimensions involved, we have a black
hole of some thirty miles in radius - in mutual orbit about a giant star
whose radius is over 300,000 times larger. Despite its small size, the gravita-
tional influence of the black hole is sufficient to distort the large star consid.-
erably out of spherical shape. It becomes rather like an egg whose small
end is somewhat pointed in the direction of the black hole. A certain amount
of material is dragged from this point and slowly falls inward to the black
hole. It does not fall straight in, however. The black hole behaves much like
a point mass. Most of the material dragged from the large star will remain
circulating about the black hole for a long time. Only gradually, as frictional
effects begin to play their part, will the material begin to spiral inwards.
Again we must bear in mind the small size of the hole. (Imagine having to
drain a normal-sized bath through a plughole a ten-thousandth of an inch
across - or a bath the size of Loch Lomond through a normal-sized plug-
hole!) The material can be only very slowly funnelled into the black hole.
And as it gets funnelled in it gets compressed and very hot - so hot that the
material must be expected to radiate light of very short wavelength, X-rays,
in fact. Such X-rays are actually observed coming from the vicinity of
HOE 226868 . And the source of these X-rays (referred to as Cygnus X-1)
appears , on the basis of detailed observations, to be in orbit about the visible
component HDE 226868 . The observed signals seem to be perfectly con-
sistent with the black-hole picture I have presented . However we should
remain cautious about drawing premature conclusions, as it is still conceiv-
able that some alternative explanation of the observations may eventually
36 Roger Penrose

turn out to be correct. The present evidence seems to be pointing ever

more strongly in favour of Cygnus X-1 being a black hole, but even if for
some reason this interpretation does turn out to be erroneous after all,
it would still be very surprising (on the basis of present theory) if no black
holes were found to exist. To indicate why, I should explain something of
the picture that astronomers and astrophysicists have developed concerning
stellar evolution and then indicate some of the theory that lies in support
of the black hole picture that I have presented.

Stellar Evolution

Let us consider first what theory and observations tell us to expect

for the future of our sun - or of any other normal star of about the same
mass. After shining at approximately its present brightness for about 7,000
million years the sun will begin a change which will transform it beyond
recognition. According to the well-accepted theory of stellar evolution the
sun will grow to an enormous size and become, like stars such as Antares in
the constellation Scorpio, a red giant some 300 million kilometers in diameter.
By this time the planets Mercury, Venus and the earth will have been burned
away and their former orbits will lie well within the new solar surface. As it
continues to burn more and more of its nuclear fuel, the bloated sun will halt
its expansion and begin to contract - down past its present size, smaller
and smaller until it stabilises as a white dwarf star perhaps about the size
of the earth. At this stage further contraction will be impossible because the
electrons of its atoms will be packed together so closely that a law of quan-
tum mechanics known as Pauli's Exclusion Principle will come into play.
In this state the density of the solar material will be such that a matchbox
full of it would weigh several tons. No material on earth has a density re-
motely approaching that of a white dwarf but as with red giants, many
white dwarfs can be seen in our galaxy. Their ultimate fate is simply to cool
off to form black dwarfs and thereafter act merely as centres of strong
gravitational attraction. The planets, Jupiter, Saturn, Uranus, Neptune, Pluto
and possibly Mars, will still continue to circle the ancient sun aeons after
it has died.
White dwarfs are part of the normal evolutionary history of average-
sized stars like the sun, and astronomical observations show actual stars
at each stage of stellar evolution through the stage now reached by our sun,
on to the red giant phase and back to white dwarfs. Moreover, the theory of
stellar physics fits these observations closely.
But not all stars can follow this 'normal' evolutionary path. As long ago
as 1931, Subrahmanyan Chandrasekhar calculated that there must be a
maximum mass above which a white dwarf could not sustain itself against
even further gravitational contraction despite Pauli's Exclusion Principle.
Black Holes 37

Many stars have masses considerably more than 1.5 times that of the sun.
What is going to happen to them?
The answer depends on just how heavy the star is. Consider a star of
twice the mass of the sun. Like the sun it will also expand to an enormous
size and then recontract, but being more massive than Chandrasekhar's
limit for a white dwarf it will be unable to settle down to final equilibrium
in the white dwa rf state. To picture what happens it will be useful to consi-
der the giant phase of a star more fully. As soon as the central density of the
star reaches that of a white dwarf, the outer layers of the star expand, and
they go on expanding as more and more of the central material gets com-
pressed into a white dwarf state. So the giant star develops a growing white
dwarf core. In the case of the sun, all the material that remains in the star
will eventually become part of this white dwarf. But if the star is too massive,
there comes a point at which the core effectively exceeds Chandrasekhar's
limit, whereupon it promptly collapses. In the process of collapse there is a
tremendous release of energy. much of which is in the form of neutrinos
which are absorbed (so it is believed) in the outer regions of the star, heating
the envelope to an enormous temperature. A cataclysmic explosion ensues -
a supernova explosion which blows off a considerable proportion of the
mass of the star. But especially interesting is the collapsed remnant of the
star left behind at the centre of the rapidly expanding cloud of ejected gases.
This core is much too compressed to form a white dwarf and it can only find
equilibrium as a neutron star.
A neutron star is tiny even by comparison with a white dwarf. A neu-
tron star rna \' be only I () kilometres in radius or only about one seven-hun-
dredth the radius of a white dwarf. The density of a neutron star could be
more than a hundred million times the already extraordinary density of a
white dwarf
A rna tehbox full of neutron star material would now weigh as much as
an asteroid a mile or so in diameter. The star's density would be comparable
with the density of the proton or neutron itself - in fact a neutron star
could in some ways be regarded as an over-sized atomic nucleus, the only
essential difference being that it is bound together by gravitation rather than
by nuclear forces. Individual awms have ceased to exist as such. The nuclei
are touching and form one continuous mass. What has happened is that the
stupendous gravitational forces have squeezed the electrons into the only
space available to them that already occupied by the protons, reversing
the usual reaction so that the star is now composed mainly of neutrons and
it is the Pauli Exclusion Principle acting on these neutrons that supplies the
effective forces preventing further collapse. This picture of a neutron star
was predicted theoretically by the Soviet physicist Lev Landau in 1932 and
the Swiss physicist F. Zwicky in 1934 and studied in detail by J. Robert
Oppenheimer, Robert Serber and George M. Volkoff in 1938 and 1939. For
years many astronomers doubted whether neutron stars could actually exist.
38 Roger Penrose

Since 1967 the observational situation has changed dramatically, because in

that year the first pulsars were observed. Since then the theory of pulsars has
developed rapidly and it now seems virtually certain that the radio and
optical impulses emitted by pulsars owe their energy and extraordinary
regularity to the presence of a rotating neutron star. At least two pulsars
reside inside supernova remnants, and this gives further support to the
theory that pulsars are in fact neutron stars.

How Black Holes Come into Being

There is a maximum mass above which a neutron star would not be

able to sustain itself against still further gravitational contraction. There is
some uncertainty as to the exact value of this maximum-mass limit. The
original value given by Oppenheimer and Volkoff in 1939 was about 0.7
times the solar mass. More recently, larger values of up to three solar masses
have been suggested. These higher values take into account the idea that the
heavy subatomic particles called hyperons may be present in addition to
ordinary neutrons and protons. But there are stars whose mass is more
than fifty times the mass of the sun. What will happen to these? It seems
exceedingly unlikely that all these stars would, as a result of their final
collapse phase, or earlier, inevitably throw off so much of their material
that their masses would always fall below the limits required for a stable
white dwarf or neutron star to be the result. In these the neutron core
would be unable to remain in equilibrium and would have to collapse further
inwards. But what other forms of condensed matter might be possible
considering that here we have densities in excess even of the fantastic value
that is maintained inside a neutron star?
In this case, theory tells us a different story: although greater densities
can be achieved, it is not possible to obtain any further stable final equili-
brium configurations. Instead the gravitational effects become so overwhelm-
ing as to dominate everything else. Newtonian gravitation theory becomes
quite inadequate to handle the problem, and instead, we must turn to
Einstein's theory of general relativity. But in so doing, we are led to a pic-
ture so strange that even the phenomenon of the neutron star must seem
commonplace by comparison. This new picture is the one which has now
earned the description of a black hole.
Briefly, a black hole is a region of space into which a star (or collec-
tion of stars or any other bodies) has fallen, but from which no light, matter
or signal of any kind can escape. Before examining this picture in some de-
tail, consider (Fig. 2) the degree of further contraction that would be neces-
sary for a neutron star to be compressed down to the size of a black hole.
A black hole of one solar mass has a radius of about 3 km, and is therefore
smaller - only by a factor of about 3 - than a neutron star of the same mass.
Black Holes 39

::::.::::

';f~~\W~

e"
Neutron
White Neutron Star
Dwarf Star
Sun

2501
t
0 100:1
~
0 700: 1 0+
{(~W:
,'"""
.
3.1 \
r-\
J
\'::',:,:,:: Bla~k
Hole

Red White
Giant ....... Dwarf

;~}1t~~(~:.
Fig. 2
The neutron star illustrated has the same mass as the sun but has a diameter 700 times
smaller than the earth.

Larger black holes are also possible, the radius of the hole being proportional
to the mass. For example, Cygnus X-I appears to be about 10 to 15 times
more massive than the sun, so its radius (assuming little rotation) is about
30 to 50 km.
The reason I emphasise the slight nature of this further contraction,
perhaps only 3 : 1, is that faced with the unsettling nature of the black holes,
people have naturally asked whether our physical theories are tenable under
these extreme conditions. But these theories seem to have worked well in de-
scribing a very large range of stars of enormously different sizes and densi-
ties. In any case, the conditions under which a black hole is formed are not
so extreme as all that- not necessarily more extreme than the situation
of a neutron star. For example, the densities involved as the collapsing star
crosses the absolute event horizon are not vastly different from those inside
a neutron star. The larger the collapsing mass, the less would be this density
- less in inverse proportion to the square of the mass. It has often been con-
sidered by astronomers that collections of mass of up to 100 million suns or
more might be involved in gravitational collapse in galactic centres and per-
haps the centre of M 87 is an example for such an object. The density at the
time such a huge mass crosses the event horizon might then be only about
that of water. So the local conditions need not be excessive when a black
hole is formed and there seems no reason to suppose that the black hole
condition might render general relativity somehow inapplicable. But also
other theories of gravitation (namely the Brans-Dicke-J ordan scalar-tensor
theory), lead to a black hole picture nearly identical to that arising in Ein-
40 Roger Penrose

stein's theory. Even in Newtonian theory a phenomenon similar to that of

a black hole may be said to occur. As long ago as 1798, Pierre Simone La-
place used Newtonian theory to predict that a sufficiently massive and
concentrated body might be invisible because the escape velocity at the
surface could be greater than the speed of light. A photon or particle of
light emitted from the surface of the body would then simply fall back to
the surface, and would not escape to be observed at large distances from the
body.
Unfortunately, however, this Newtonian argument does not bear close
scrutiny (and there is some evidence that Laplace himself had second thoughts
about it) because in Newtonian theory one would not expect the speed of
light to be a constant - indeed, a horizontal mirror at the body's surface
would reflect light falling from infinity back to infinity, so the velocity of
light at the surface would exceed that at infinity. But a corrected version of
Laplace's argument does hold in relativity theory!

Non-rotating Black Holes

To begin with, consider the standard picture of a non-rotating black

hole resulting from the theory of general relativity. The radius of the event
horizon can be calculated by
R _ 2MG
- c2
where c = 3 .10 8 mls stands for the speed of light and G = 6.67.10- 11 NM2 I
kg2 the gravitational constant. This radius is known as the Schwarz schild
radius and has the value of about 3 kilometres for the sun.
The star whose collapse is responsiDle for a black hole's existence has
fallen deep inside the event horizon. The gravitational field inside the event
horizon has become so powerful that even light itself is inevitably dragged
inward regardless of the direction in which it is emitted. Outside the event
horizon light escapes if it is aimed suitably outward. The closer the emission
point is to the event horizon, the more the wave front of the emitted signal
is displaced back toward the centre of the black hole. This applies not just
to light but to any signal originating within the black hole. As for a photon
emitted radially outward from the surface of the black hole, on the event
horizon, it will mark time, forever hovering in the surface itself at the same
distance from the centre of the black hole.
It may seem that this is odd physics indeed, quite unlike the normal
situation in relativity theory, where the speed of light has always the same
constant value in all directions. But, strange as it may seem, the local phy-
sics in the neighbourhood of the absolute event horizon is the same as else-
where. An observer at the event horizon who tries to measure the speed of
Black Holes 41

light must himself be crossing the horizon by falling inwards into the hole.
To him the speed of the light hovering on the horizon is indeed the same
constant value, in the outward direction. It would be natural for a reader
who is not familiar with general relativity theory to find such a situation
confusing. This is partly because so far we have been using a purely spatial
description rather than a space-time one - and for many purposes a space-
time picture is more illuminating than a spatial one. Strictly speaking, a
space-time picture needs to be drawn in four dimensions, but an overall
description of the space-time situation can be obtained by suppressing one
of the spatial co-ordinates in the space-time diagram and substituting a
time co-ordinate. This gives an instantaneous picture of what is going on
at all times, and obviates the need for many sequential 'snapshots' of a
developing situation.
Consider a light flash emitted in all directions from a given point in
ordinary space (Fig. 3). The wave front of the flash would be a sphere
centred on the emitting point and growing larger each moment at the speed
of light. A purely spatial representation of the flash would be a sequence
of spheres (Fig. 3) each sphere larger than the preceding one, marking

(01 ''. t,
( )
I , Space
, I
I I
I

Spot.

Fig. 3
,, (a) propagation of light from
,
"- X in space at times t I
", and t2'
(b) propagation of light in
(bl space-time to give a 'ligh t
cone',
42 Roger Penrose

the position of the light flash's spherical wave front at a given moment in
time. A space-time representation of the light flash, however, would be a
cone whose vertex represents the time and place at which the light flash
is emitted, the cone itself describing the history of the light flash.
By the same token the history of a star's collapse down to a black hole
can best be depicted in a space-time representation (Fig. 4). The locations
of the light cones at various points in space-time show how light signals pro-
pagate in the gravitational field. At some points the light cones are drawn
as being tipped over, but this is not something that would be noticed by a
local observer. Such an observer would follow a path in space-time that pro-
ceeds into the interior of the light cone; his speed can never be greater than
the locally measured speed of light, and only inside the light cone is this
criterion met. But the tipping of the light cones does affect what an observer
at large distances can see. Fig.4 shows that material particles and light
signals which originate inside the event horizon are inevitably driven further
inwards. For a particle or signal to cross the event horizon from the inside
to the outside it would have to violate the condition mentioned above; it
would have to exceed the local light speed, which would be inconsistent
with relativity.
By taking a horizontal section through the space-time picture we get a
spatial representation of the situation as in Fig. 5. Perhaps the main advan-
tage of a spatial representation, apart from its greater familiarity, is that
one need not dispense with the third spatial dimension when it is important
not to do so. The light cones can be depicted as points (the origin of a light
flash) surrounded by sphere-like surfaces (the location of the light flash a
moment later) - except that when the light cone is tipped over, these sur-
faces do not actually surround the point of origin. In this later case it would
be necessary to exceed the local light speed in order to 'stay in the same
place'. A serious drawback with such spatial pictures, therefore, is that it
becomes hard to interpret situations of this kind. If a space-time description
is used, then it becomes easier to accept that i:he local physics is the same
whether or not the light cones are depicted as tipped over, the 'tipping'
being merely an aspect of the local description.
Although nothing can ever get out of a black hole, things can fall in.
Indeed, it is quite possible that stellar astronauts traversing the depths of
space in ages to come will run precisely this risk. Not that they will be
likely to encounter a black hole by accident - the smallness of black holes
compared with the vastness of the universe will see to that. Indeed, they
would have to seek out a black hole deliberately if they wished to expe-
rience this 'ultimate trip'. And what will happen to a hapless astronaut
who falls into a black hole? What, indeed is the fate of the original body
which collapsed to produce the black hole? Assuming that the exact spheri-
cal symmetry is maintained right down to the centre, the answer provided
by general relativity is an alarming one. According to the theory, the curva-
Black Holes 43

l ime Singulorlt_Y:'---it-_~_ EYent Horizon

, «q9
~
~

~ ~
~ ~
~
~
~ ~
Empty space Empty SPQce
Fig. 4
TIme
Space-time representation
~pace ~
Collapsing
maUer ~ of collapse of a spherically
SpQce symmetrical body to form
a black hole, showing
propagation of light inside
~ and outside the event
------------ horizon.
'.
"

®
®
®
/' """0
Event Honzon
- , ,,
.... ®
I x
p 0 \

®
I
\
\
,, 0
)(
•
Singularity /
I
I

® ®
--- ""
"- ...... ./

Fig. 5
Light propagation inside and outside a non-rotating black hole.
44 Roger Penrose

ture of space-time increases without limit as the centre is approached.

Not only is the material of the original body squeezed to infinite density at
the centre of the black hole - and crushed, effectively, out of existence -
but also the vacuum in space-time which is left behind by the body, itself
becomes infinitely curved. The effect of this infinite curvature on an observer,
were he foolhardy enough to follow the body inwards, would be that he
experiences mounting gravitational tidal forces - tidal forces that mount
rapidly to infinity.
The gravitational tidal effect is the most direct physical manifestation
of space-time curvature. Einstein pointed out that the gravitational force on
a body can be eliminated at anyone point simply by choosing a frame of
reference that is falling freely. He gave the famous example of a lift that
broke its cable and fell toward the earth. Any passenger inside would be
falling at the same rate as the lift, so he would feel no net gravitational force
relative to the lift, and, indeed, would float free of gravity inside it. Such
elimination of the gravitational force by free fall is now a familiar feature
of space travel. The tidal effect, however, cannot be so eliminated and is,
therefore, an absolute manifestation of the gravitational field. Imagine an
observer falling freely in the earth's field. Suppose he is surrounded by a
sphere of particles that he initially observed to be at rest with respect to
himself. The Newtonian gravitational field of the earth varies as the inverse
square of the distance between it and any other body, pulling more strongly
on objects closer to its surface than on objects farther away. This non-
uniformity of the gravitational field tidally distorts the sphere of particles
into an ellipsoid. The earth's ocean tides are a familiar example of an effect
of this kind.
Fortunately for us, the tidal effects due to gravity encountered in the
solar system are small. Nobody complains that his feet experience more
gravitational pull towards the earth than his head. There is a difference,
but it is not noticeable in the ordinary way. The space-time curvature
responsible for this difference has a radius of about the distance from the
earth to the sun. This is a pure coincidence since the sun is itself irrelevant
to this particular tidal effect. At the surface of a white dwarf, on the other
hand, the space-time curvature is considerably larger, the radius of curvature
being of the same order as the radius of the sun. This tidal effect would be
very noticeable to an astronaut in orbit around the white dwarf. In fact, his
head and feet could experience a difference in forces of perhaps one-fifth
of the total force the astronaut normally experiences standing on the earth's
surface. At the surface of a neutron star, however, the tidal effect is, by
ordinary standards, enormous. The radius of space-time curvature there is
only about 50 km. It is clear that no astronaut in a low orbit around a neu-
tron star could possibly survive. For even if he curled himself into a small
ball, the gravitational acceleration at various parts of his body would differ
by several million times the gravity at the earth's surface. However, instru-
Black Holes 45

ments could in principle be built to withstand such high tidal forces - all
that would be necessary would be to make them tiny enough.
Suppose our astronaut carries such a tiny, rugged instrument as he flies
towards a black hole of one solar mass. Long before he reaches the event
horizon he would be destroyed by the tidal forces, but his instrument can
survive intact as it crosses the event horizon, where it experiences tidal
forces at least thirty times those at the surface of the neutron star. As the
instrument falls in towards the centre, the mounting tidal forces will rise
rapidly, ripping to pieces in turn the material of the instrument, the mole-
cules of which this material is composed, the atoms which constitute these
molecules, the atomic nuclei, and finally the fundamental particles which a
moment ago had been the building-blocks of these nuclei. And the entire
process would not last more than a few thousandths of a second!
So anything falling into a black hole, whether it be a space ship, a hy-
drogen molecule, an electron, radio-waves or a beam of light can never
emerge again. So far as our universe is concerned it disappears completely
and forever into nothing. But how can this be? Is it not a basic law of
nature that matter or energy can never be completely destroyed but only
converted from one form into another? The question is a perfectly respect-
able one, but it can be shown by rigorous argument, based on general relativ-
ity, that there must be a region inside a black hole, a region of infinite
curvature, called a space-time singularity at which the known laws of physics
break down. So there is no known conservation law that can be relied on at
the centre of the black hole. Eventually, perhaps, laws of nature may be
formulated which govern the behaviour of space-time singularities, but no
such laws are known at present. General relativity, like virtually all viable
physical theories, is reversible in time. So corresponding to any solution of
the questions in which time runs one way, there must be another in which
the time-sense is reversed. This leads us to expect that the above situation
could - in principle .- exist in a time-reversed form. Initially there would be
the space-time singularity. Then matter would appear: elementary particles,
light. Only later would these particles collect together into atoms, molecules
or stars. In fact, a picture of this kind has been considered for many years as
a model of the creation of the universe.
The initial big bang of the cosmological models is, like the centre of a
black hole, also a space-time singularity, where the curvature of space-time
becomes infinite. But now, rather than being destroyed, matter is created
at the singularity. The cosmological big bang is not precisely the time-reverse
of a black hole, however, since the singularity is all-embracing, unlike the
relatively localised singularity inside the black hole. The basic difference is
one of size, and we may indeed envisage more localised 'little bangs', called
white holes, which are more precisely the time-reverses of black holes.
A number of theoreticians have considered such white holes seriously in
connection, in particular, with models for quasars. However, I must say that
46 Roger Penrose

I personally regard the possibility of the existence of white holes with consid-
erable unease - and disbelief! The reason is basically this. As soon as a black
hole is formed there is (within classical physics) no means of destroying it. It
is created violently, but then settles down and sits around forever - or until
the universe re-collapses at the end of time. Now a white hole - the time
reverse of a black hole - would have had to have been there since the begin-
ning of time - tamely and invisibly biding its time before making its presence
known to us. Then, when its moment arrives, it explodes into ordinary matter.
But this moment is of its 'own' choosing, governed, apparently, by no def-
inite law. As yet, there is no clearly understood and definitive theoretical
argument against white holes. But nevertheless, certain arguments can be
given to show that their presence would be totally at variance with certain
very desirable thermodynamic principles. In my opinion, we do not have to
take them seriously as actual objects in the universe. However, we are still
stuck with the big bang and that perhaps seems untidy. But here there
appears to be no way out.
But let us now return from these questions and pursue the argument
concerning black holes. Quite apart from the doubts I have already raised
about the validity of the general theory of relativity, there are other questions
that need to be settled before one can fully accept the theoretical concept of
the black hole as a realistic description of something that actually occurs in
nature. In the first place, can we be sure that enough is known about the
behaviour of matter under the extreme conditions required to form a black
hole for the predictions to carry conviction? What role does the assumption
of exact spherical sy~metry play in the discussion? Let us consider these
questions in turn.
As I have already pointed out, the densities involved in the formation
of a black hole need not be excessive. The same applies to space-time curva-
tures. Especially where black holes with a mass equal to several million solar
masses are concerned, an astronaut passing through the absolute event hori-
zon would notice nothing. He would have no means of telling that an irre-
trievable situation had developed, because the exact location of the horizon
is not something that can be discerned by local measurement. After this
he would have but a few hours to enjoy the experience of life inside a black
hole before the tidal effects mounted to infinity.

Asymmetrical Collapse

The question concerning the role played by the assumption of spheri-

cal symmetry has to be examined more carefully. If we do not assume
spherical symmetry, then we cannot appeal to the exact solutions of Ein-
stein's equations on which we have based the foregoing discussion. Further-
more, even if we assume that initially the deviations from spherical symmetry
Black Holes 47

are slight, we should have every reason to expect that near the central point
these asymmetries would be enormously magnified. Might not the different
portions of the collapsing body miss one another? Perhaps they could re-
emerge after a close encounter and bounce out again. It is fortunate that,
owing to some general theorems that have been proved over the past several
years, a remarkably complete picture of asymmetrical collapse has emerged.
Considering the picture in a little detail, suppose that a massive star or
a collection of bodies collapses and that deviations from the spherical sym-
metry are at first comparatively small. We can establish that a point of no
return has been passed if a certain criterion is satisfied. Imagine that a flash
of light is emitted at some instant at some point in space. The flash of light
will follow the light cone centred on the point according to our space-time
representation (Fig. 6). The light rays start out from the point by diverging
in all directions. When they pass through matter or through a gravitational
field, the matter or the field has a focusing effect on the rays. If enough
matter or a sufficiently strong gravitational field is encountered, the amount
that the rays diverge can be reduced to such an extent that this divergence
is actually reversed, that is, the rays start to converge. The required criterion

Singularity

Infalling ma ter ial

,Tim~
Fig. 6
IL;:ace The birth of a black hole.
Space
48 Roger Penrose

for a point of no return is that every light ray from the space time point
encounter enough matter or gravitation for the light cone to be reconverged.
It is not hard to show from simple order-of-magnitude estimates that, for
sufficiently large collections of mass, the criterion can indeed be satisfied
before densities or curvatures became excessive, and without any assump-
tion of symmetry. Once this criterion has been satisfied, then according
to a precise theorem in general relativity put forward by Stephen Hawking
and myself, it follows that there must be a space time singularity somewhere.
The theorem does not say that this singularity is necessarily of the same
character as that encountered in the centre of a spherically symmetrical
black hole, but it is hard to avoid the inference that tidal effects which ap-
proach infinity will occur, producing a region of space-time where infinitely
strong gravitational forces literally squeeze matter and photons out of
existence.
Physicists are unhappy with a theory that predicts the evolution of such
a truly physical singular state. In the past whenever a singularity was en-
countered in a theory, it was generally a warning that the theory in its pres-
ent form was breaking down and new theoretical tools were needed. In the
case of black holes we theoreticians are again being presented with a situa-
tion of this kind, but one more serious than before, because here the singu-
larity refers to the very structure of space and time.
There are two distinct possibilities at this stage. It may be that the
resulting singularity is such that signals can escape from it which can be
observed at large distances. Such a singularity is called 'naked'. The possi-
bility of naked singularities is alarming because the physical effects of
near-infinite space time curvatures are quite unknown. If these effects can
influence the outside world, then an essential uncertainty is introduced
into present physical theory.
On the other hand it is possible that the singularities resulting from
gravitational collapse are always hidden from view, as was the case in the
spherically symmetrical situation considered above. This is the hypothesis
of 'Cosmic Censorship', according to which naked singularities are for-
bidden, each singularity being necessarily clothed by an absolute event
horizon. There is perhaps some slight theoretical evidence in favour of
Cosmic Censorship, but it is only rather slight. However, it would be in-
accurate to think of the big bang as a violation of Cosmic Censorship. We
are concerned here only with singularities which arise in the collapse of per-
fectly reasonable non-singular matter. I would certainly tend to believe
that in situations which do not differ too much initially from that of spheri-
cal symmetry, the Cosmic Censorship principle is valid. In more extreme
cases the question is much more open. My own opinion has come round to
the view that Cosmic Censorship is likely always to be valid (in situations of
classical relativity).
Black Holes 49

If we assume the Cosmic Censorship hypothesis is true, then once the

focusing criterion is satisfied an absolute event horizon must arise. This
horizon will have a well-defined cross sectional area which will have a tend-
ency to increase with time (black holes can grow but, classically, can never
shrink) but it seems reasonable to suppose that a black hole, left to itself, will
settle down to a stationary state. However, we must be cautious in our use of
intuition. It might also seem reasonable to suppose that given the vast
range of structures, configurations and complexities of the bodies which
could have collapsed to a black hole in the first place, the configuration
of the black hole itself could also be complex. Some remarkable work by
Werner Israel, Brandon Carter, Stephen Hawking, and David Robinson
has shown that this is not the case. Only a very restricted class of stationary
black hole configurations can arise. They are uniquely characterised by the
value of the mass, spin and charge of the hole. Einstein's equations for the
general theory of relativity have been solved explicitly for this problem by
Roy Kerr and the solution generalised to include charge by Ezra Newman
and his co-workers. The reason the asymmetries present in the collapsing
body do not show up in the final state of the black hole is that once the hole
is formed the body that produced it has little influence on the hole's sub-
sequent behaviour. The black hole is best thought of as a self-sustaining
gravitational field governed by the internal non-linear dynamics of the general
theory of relativity. These dynamics allow the asymmetries in the gravitational
field of the hole to be carried away in the form of gravitational waves as the
hole settles down into a stable configuration.

Black Holes as Energy Sources

We have seen that a material object, once swallowed by a black hole,

cannot escape. On the other hand, there are mechanisms whereby some of
the energy content of the black hole can be extracted. One such mechanism
involves the coalescing of two black holes. This process would be accom-
panied by the copious emission of gravitational waves, whose total energy
should be a substantial fraction of the initial rest mass energy of the black
holes. Another mechanism would be to allow a particle to fall into a region
close to the event horizon of a rotating black hole. The particle can be made
to split into two particles in such a way that one falls into the hole and the
other escapes back to infinity with more mass energy than the initial particle
had. In this way rotational energy of the black hole is transferred to the
particle motions outside the hole. In the process the black hole loses mass
and spin. In principle, this is an extremely efficient means of converting rest-
mass to energy - much more efficient than nuclear fission or fusion! In the
most extreme case the mass of the black hole might conceivably be reduced
down to 0.707 of its original value by this sort of general procedure. But it is
hard to envisage this process being effective in most astrophysical situations.
50 Roger Penrose

Black Holes and the Quantum Theory

Let us now consider the situation inside the black hole, and the general
relativistic implication of the existence of a space-time singularity. Since a
'singularity' means a region of breakdown of physical theory, we have the
curious situation that, here, general relativity is predicting its own downfall!
But perhaps we should not be too surprised at this; after all we are treating
general relativity only in its capacity as a classical theory. When the curva-
ture of space time becomes enormous, quantum effects must eventually
playa dominant role. When the radius of space-time curvature becomes as
small, as say, (10- 15 ) m (roughly the radius of an elementary particle) then
the theory of particle physics as understood at present, must break down.
If the radius of space-time curvature ever becomes as small as (10- 35 ) m
(and the implications of what we have said so far are that it will be that
small somewhere inside a black hole - unless theory breaks down before
this), then we cannot avoid having to apply quantum mechanics to the
structure of space-time itself. At present there is no satisfactory theory for
doing this. And if Cosmic Censorship holds true, the absolute event horizon
would prevent the effects of quantum physics in the centres of black holes
having any influence on the outside world.
But one of the most important theoretical developments of the past
several years has shown that our previous considerations do not hold com-
pletely true. With very tiny black holes, whose Schwarzschild radius is approxi-
mately 10- 15 m (the radius of an elementary particle) quantum effects be-
come important, which lead to a destruction of the black hole in a big explo-
sion. This is the discovery by Stephen Hawking, following some results of
Jacob Bekenstein, that if quantum effects are taken into account, a black
hole should continually radiate a certain amount of energy. For any of the
holes that could be expected to arise in astrophysical processes, such as in
Stellar collapse, this effect is utterly minute and does not affect the previous
classical discussion. But it is just conceivable that very tiny holes (of, say,
the diameter of an elementary particle) might have been produced in a
highly chaotic big bang. For such holes the radiation would be very large
and would increase in intensity as the hole loses its mass-energy in radia-
tion. The final explosion of such a mini hole could be detectable here on
earth. However, the observational evidence is such as to indicate that mini-
holes, if they exist at all, must be very infrequent indeed (and there are also
some theoretical arguments against them). Nevertheless, the theoretical
implications of Hawking's result are very considerable, as they suggest
hitherto unsuspected interrelations between general relativity, thermodynam-
ics and quantum field theory.
 51

The Black Hole:

An Imaginary Conversation with Albert Einstein.
John A. Wheeler

Wheeler. Professor Einstein? Professor Einstein! Oh how wonderful to see

you here!
Einstein: Yes, I thought I would surprise you Mr. Wheeler. But you know
I like the sea shore and the waves as much as you do.
Wheeler: But the biggest miracle of all is not our picking the same prospect
over sea and distant islands, it is that you should be here at all.
What wonderful fortune!
Einstein: Yes, I thought you would be surprised. I am too. Could it have
anything to do with those old stories that one is permitted to
come back to Earth again for a single hour on one's hundredth
birthday? Do you remember Niels Bohr's story about why hang
up a horseshoe over your desk?
Wheeler: No, I don't think I heard that one.
Einstein: He used to say, "I don't believe in miracles at all- and especially
not in horseshoes. But you know that people who do tell me
that it doesn't hurt your luck if you don't believe." [1] That must
be the general idea of how come I am here now.
Wheeler: Less than a hour left and that to rejoice in all this beauty. No won-
der you don't want to ask questions but only look and smile and
close your eyes and look again. But while you look could I ask
you some questions? So many colleagues have regretted with me
all the questions, great and important questions, that we failed to
ask you before you left our midst.
Einstein: But yes. Still you must not consider me an expert. When I was
young I made so much trouble for authority that in later life, in
punishment, the Lord made me an authority myself [2]; but I
am no more.
Wheeler. Why did you not say more about what we today call the "black
hole "?
Einstein: Yes, I know what you mean: a completely collapsed star. It was
not so easy to discuss such questions in my time. What after all
is one to take for the equation of state of the matter of the star?
That is not so easy.
52 John A. Wheeler

Wheeler: Then I'm sure you'll rejOICe in the theorems of today that a
sufficiently massive collection of cold matter has no escape from
gravitational collapse.
Einstein: Of course it is much simpler to stay away from all special assump-
tions about the relation between density and pressure. That is why
in my 1939 paper [3] I considered a collection of well separated
point masses in orbit about their mutual center of gravitation.
One discovers that the cluster of mutually gravitating particles is
unstable.
Wheeler: Stimulating that was to us all as a beginning. The future will surely
see that work continued from the initial stages of instability to the
final stages of collapse. But did you never feel an urgency about
an extension of these ideas to real stars and an astrophysical search
for real black holes?
Einstein: Yes, that is an interesting point. But, no, for me that was an un-
important question of detail. One has only to apply the general
relativity theory itself to come to some somewhat reliable conclu-
sions about collapse.
Wheeler: How do you feel about Kerr's exact solution for the geometry
around a rotating black hole, consequences of Kerr's and Schild's
early mathematical investigations of algebraically special solutions
of your gravitational field equation?
Einstein: That is really beautiful! I would never have dared to hope for an
exact and simple solution of a problem so difficult.
Wheeler: And how do you react to the work of Carter [4,5] and others?
Einstein: How wonderful it is that they can show that the geometry around
a collapsing object, no matter how contorted and asymmetric and
violently fluctuating it is, in the end always tends smoothly to an
absolutely standard final state depending only on the mass and
electric charge and angular momentum of the black hole. Of course,
that is outside. But, after all, the critical place is inside. That's
where the predicted singularity is. That's where the problem is.
It is impossible to believe a prediction that is a singularity.
Wheeler: I and my colleagues have to confess that we have made only a bare
beginning at studying the approach to singularity both in cosmolo-
gy and in black hole physics.
Einstein: To understand that approach is really important.
Wheeler: Our Soviet colleagues propose fascinating physical insights as to
what goes on in the final stages of collapse, but not convincing
mathematical methodology. Colleagues in the West have the
mathematical methodology but so far it has not sufficed to pro-
vide the insight that we all want.
Einstein: This is an old story in physics. We know in the end everything
comes together in a new and better and larger unity.
The Black Hole: An Imaginary Conversation with Albert Einstein 53

Wheeler: Gamow says that you called it "the greatest blunder of my life" [6).
The 1929 observation of Hubble [7) revealed that you need not
have added an artificial "cosmological term" to your standard
1915 general relativity theory to secure a reasonable account of
cosmology. Also it was Friedmann [8), not you, who in 1922 first
worked out that even today standard and simple cosmology
from your field equation. I and many colleagues want to ask
you how this came about. Was it because [9] you had taken
very seriously in your younger days the idea of Spinoza that the
universe endures from everlasting to everlasting [1O]? Did any
contrary conclusion seem to you philosophically unreasonable? In
your autobiographical notes [11] you speak of the influence of
Spinoza on your outlook. It has been said that "Spinoza rejected
the idea of an external Creator suddenly and apparently capri-
ciously creating the world at one particular time rather than
another and creating it out of nothing." [10] Do you feel that
Spinoza indeed so powerfully influenced what you did and did not
do in the field of cosmology?
Einstein: That is very hard to say. Even today we do not know the right
way to think about these questions.
Wheeler: Do you envisage a "before" before the big bang and an "after"
after the big crunch?
Einstein: One can hold many views on these questions and they all deserve
consideration.
Wheeler: This is an exciting time in astrophysics. Some colleagues are in-
clined to believe that the universe contains less than a tenth of
the mass energy that would be required to curve it up into closure.
Others, following the lead of other evidence, find indications that
the amount of mass energy may be close to what you predicted.
What is your view on the question whether the universe is closed?
Einstein: "[We] may present the following arguments against the concep-
tion of a space-infinite, and for the conception of a space-bounded,
universe: (1) From the standpoint of the theory of relativity, the
condition for a closed surface is very much simpler than the cor-
responding boundary condition at infinity of the quasi-Euclidean
structure of the universe. (2) The idea that Mach expressed, that
inertia depends upon the mutual action of bodies, is contained,
to a first approximation, in the equations of the theory of rela-
tivity; ... But this idea of Mach's corresponds only to a finite
universe, bounded in space, and not to a quasi-Euclidean infinite
universe." [12] "In my opinion the general theory of relativity
can only solve this problem [of inertia] satisfactorily if it regards
the world as spatially self-enclosed." [13]
54 John A. Wheeler

Wheeler: I don't have to tell you that there is still a non-negligible body
of our colleagues who think that an asymptotically flat universe
is more natural than a closed universe.
Einstein: But that view takes the geometry of faraway space out of physics
and makes it pan of theology, to be discovered by reading Euclid's
bible. It puts us back to the days before Riemann, days when
space was still for physicists, "a rigid homogeneous something,
susceptible of no change or conditions. Only the genius of Rie-
mann, solitary and uncomprehended, had already won its way
by the middle of the last century to a new conception of space,
in which space was deprived of its rigidity, and in which its power
to take part in physical events was recognized as possible. " [14]
Wheeler: Space, a new participant in dynamics - that's what you gave us
in your equations! Elie Canan recognized that your geometro-
dynamics requires initial value data just as does any other dynam-
ics [15, 16]. How did you react to his investigations?
Einstein: Yes, Cartan saw into the mathematics deeper than anyone. Yes,
I recognized the importance of his work. Yes, I told Helen Dukas,
"Don't file Cartan's papers away as you do other papers; keep
them out separate so I can study them." Canan understood
things more clearly than anyone.
Still there are two sides to the initial value problem. One is their
consequences. They we know in a general way how to calculate.
The other is their origin. We still have not the faintest idea of
what considerations fix the initial conditions. Your Peebles at
Princeton and his colleagues [17] have studied the initial condi-
tions for cosmology more fully than anyone. They show that
things in the beginning were not quite so arbitrary as one might
have thought. They also find indications that the density is of
the same order as what the general theory predicts. That result
seems to me a natural one.
Wheeler: Do you then feel that gravitational collapse of the universe is
similar in principle to gravitational collapse of a star to a black
hole?
Einstein: To think of both processes as equally inescapable is reasonable.
I confess it was a surprise to me about the big bang. But once
we have to accept that, it seems to me only consistent that we
should also accept gravitational collapse, both for stars and for
the universe. Yes, that is quite contrary to the idea that Spinoza
taught that the universe goes on forever. You mentioned his argu-
ment against an original creation. How could nothingness, depriv-
ed of all possibility of knowing time, know when to give birth
to the universe? How are we to answer this objection today?
The Black Hole: An Imaginary Conversation with Albert Einstein 55

It must be that time is not so primordial a concept as we take

it to be. It must first then come into being when the universe
itself begins. That diminished status for time may not be so
unreasonable.
Wheeler: But if time is not truly basic, can geometry itself be truly basic?
And in that case what happens to your vision of all the forces
of nature taking their origin, one way or another, in geometry?
Einstein: The workers of today have a wider understanding of what geo-
metry is and means than was current in my lifetime. Gauge theo-
ries - what are they but a new and deeper version of geometry?
Even spinor fields nowadays have the "geometry hat" clapped
upon their heads, I am told.
Wheeler: But whether you call particles geometry or something else, does
it not trouble you that collapse should mean their end?
Einstein: To me the problem of collapse is no greater than the problem
of the big bang. Both are a warning that the universe presents
deeper issues than we ever realized. That to me is the lesson
of the black hole. Alas, I can say no more. I feel myself being
carried away, not to return for another hundred years. But let
me leave you hope for the work of all your colleagues. "All
of these endeavors are based on the belief that existence should
have a completely harmonious structure. Today we have less
ground than ever before for allowing ourselves to be forced
away from this wonderful belief." [18]

References

[1] Weber, R. L., "A Random Walk in Science", (Institute of Physics, London) 1973
p. 14. I.B. Cohen -+ S. Gouldsmit -+ N. Bohr.
[2] Einstein, A, As quoted in B. Hoffmann, "Albert Einstein: Creator and Rebel",
(Viking, New York) 1972. p. 24.
[3] Einstein, A, "On a stationary system with spherical symmetry consisting of many
gravitating masses", Ann. Math. (USA) 40, 922-936 (1939).
[4] Carter, B., "An axisymmetric black hole has only two degrees of freedom", Phys.
Rev. Lett. 26,331 333. (1970).
[5] Carter, B., "Properties of the Kerr metric", in "Black Holes", Proceedings of 1972
sessions of Ecole d'ete de physique theorique, C. De Witt and B. S. De Witt, eds.,
(Gordon and Breach, New York) 1973.
[6] Gamow, G., "My World Line" (Viking Press, New York) 1970.
[7] Hubble, E. P., "A relation between distance and radial velocity among extragalactic
nebulae." Proc. Nat. Acad. Sci. U.S. 15, 169-173 (1929).
[8] Friedmann. A, "Uber die Kriimmung des Raumes." Z. Phys. 10, 377-386.
[9] The importance of Spinoza's philosophy for Einstein's outlook was kindly emphas-
ized to me by Hans Kung at Tiibingen 12 June 1978.
56 John A. Wheeler

[10] Wolf, A., "Spinoza", "Encyclopedia Britanica", Chicago, 1956, Vo!' 21, p. 235.
[11] Schilpp, P. A., ed. "Albert Einstein: Philosopher-Scientist", (Library of Living Phil-
osophers, Evanston, Ill.) 1949.
[12] Einstein, A., "The Meaning of Relativity", 3rd edition, (Princeton University Press,
Princeton, N.].) 1950. pp. 107-108.
[13] Einstein, A., "Essays in Science", (Philosophical Library, New York) 1934. Trans-
lated from "Mein Weltbild", (Querido Verlag, Amsterdam) 1933. p. 55.
[14] ibid. p. 68.
[15] Cartan, E., "Sur les equations de la gravitation de Einstein", ]. Math. Pures App!. 1,
141-203 (1922).
[16] Cartan, E., "La theorie des groupes et les recherches recentes de geometrie differen-
tielle", "Conference Proceedings International Congress of Mathematicians", To-
ronto, (1924) L'Enseign. math. t. 24, 1-18 (1925).
[17] Davis, M., E. J. Groth und P. J. E. Peebles, "Study of galaxy correlations: Evidence
for the gravitational instability picture in a dense universe", Astrphys. ]. 212:
Ll07-Ll11 (1977).
[18] Einstein, A., "Essays in Science", (Philosophical Library, New York) 1934. Trans-
lated from "Mein Weltbild", (querido Verlag, Amsterdam) 1933. p. 114.
 57

Can QIantum-Mechanical Description of Physical

Reality Be Considered Complete?
Nathan Rosen

Introduction

In 1935 there appeared a paper with the above title (Einstein, Podolsky
and Rosen 1935) in The Physical Review. This paper was the outgrowth of a
number of discussions held by Albert Einstein, Boris Podolsky and myself
at the Institute for Advanced Study in Princeton. The purpose of the discus-
sions was to help us understand the concepts and principles of quantum
mechanics, and what we thought we understood troubled us. For the con-
clusion we reached in these discussions was that the answer to the above
question is, "No'.
The paper aroused considerable controversy among physicists at the
time. Now, more than forty years later, discussion is still going on. It seems
therefore appropriate on the occasion of the centennial of the birth of Albert
Einstein to go back to this paper and re-examine it from the perspective of
the present time.
The next section presents a detailed review of the above paper (to be
referred to hereafter simply as "the paper") together with some critical
remarks. This is followed by a section presenting a contrasting point of view,
that of Bohr, and then a section of discussion.

II The Paper

The paper begins with the statement:

Any serious consideration of a physical theory must take into account the distinc-
tion between the objective reality, which is independent of any theory, and the physical
concepts with which the theory operates. These concepts are intended to correspond with
the objective reality, and by means of these concepts we picture this reality to ourselves.

Obviously, it is tacitly assumed, as most physicists believe, that there

exists an objective reality, a physical world independent of the human ob-
server, and that a physical theory describes some aspects of this reality,
thereby enabling us to form some sort of picture of it.
58 Nathan Rosen

In attempting to judge the success of a physical theory, we may ask ourselves two
questions: (1) "Is the theory correct?" and (2) "Is the description given by the theory
complete?" It is only in the case in which positive answers may be given to both of these
questions, that the concepts of the theory may be said to be satisfactory. The correctness
of the theory is judged by the degree of agreement between the conclusions of the theory
and human experience. This experience, which alone enables us to make inferences about
reality, in physics takes the form of experiment and measurement. It is the second ques-
tion that we wish to consider here, as applied to quantum mechanics.

Nowadays one sometimes hears it said that all that one wants of a theory
is that it should be correct, i. e., that it should enable one to carry out calcula-
tions so as to obtain numbers that agree with the results of experiments, and
that it is not necessary for the theory to provide us with any picture of the
reality. However, it seems to me that most physicists want such a picture,
and for them the second question, that of completeness, is important.
Whatever the meaning assigned to the term complete, the following requirement
for a complete theory seems to be a necessary one: every element of the physical reality
must have a counterpart in the physical theory. We shall call this the condition of com-
pleteness. The second question is thus easily answered, as soon as we are able to decide
what are the elements of the physical reality.

It might be remarked that, in addition to deciding what is an element

of the physical reality, one also has to decide what is its counterpart in the
theory. It is taken as self-evident here that the corresponding concept and
the numerical value associated with it should appear in the theory.
The elements of the physical reality cannot be determined by a priori philosophical
considerations, but must be found by an appeal to results of experiments and measure-
ments. A comprehensive definition of reality is, however, unnecessary for our purpose.
We shall be satisfied with the following criterion, which we regard as reasonable. If,
without in any way disturbing a system, we can predict with certainty (i. e., with prob-
ability equal to unity) the value of a physical quantity, then there exists an element of
physical reality corresponding to this physical quantity. It seems to us that this criterion,
while far from exhausting all possible ways of recognizing a physical reality, at least
provides us with one such way, whenever the conditions set down in it occur. Regarded
not as a necessary, but merely as a sufficient, condition of reality, this criterion is in
agreement with classical as well as quantum-mechanical ideas of reality.

This criterion is of crucial importance to the discussion. The key point

in it is: "without in any way disturbing the system". This will be considered
below.
To illustrate the ideas involved the paper considers the quantum-
mechanical description of the behavior of a particle having one degree of free-
dom. Reference is made to the concept of state, completely characterized by
the wave function l/J, to the correspondence between each physically observ-
able quantity A and an operator (also denoted by A), and to the eigenfunc-
Quantum-Mechanical Description of Physical Reality 59

tions and eigenvalues of this operator. It is recalled that, if l/J is an eigenfunc-

tion of the operator A, 1. e.,
A l/J = a '1/1, (1)

where a is a number, then the observable A has with certainty the value a
whenever the particle is in the state given by '1/1.

In accordance with our criterion of reality, for a particle in the state given by t/J
for which Eq. (1) holds, there is an element of physical reality corresponding to the
physical quantity A.

As an example, the state is considered for which '1/1 is given by

'1/1 = e(2rrilh)pox, (2)

where Po is a constant and x the independent variable. With the momentum

operator P = (h/21f i) a/ax, one sees that l/J is a momentum eigenfunction
with eigenvalue Po.

Thus, in the state given by Eq. (2), the momentum has certainly the value Po. It
thus has meaning to say that the momentum of the particle in the state given by Eq. (2)
is real.

On the other hand, if one considers the particle coordinate, for which
the operator q is that of multiplication by x, Eq. (1) does not hold in this
case, and one cannot say that the coordinate has a particular value. In accord-
ance with quantum mechanics one can talk only about probabilities of
various values, and in the case of the state given by Eq. (2), it is shown that
all values of the coordinate are equally probable.
A definite value of the coordinate, for a particle in the state given by Eq. (2), is
thus not predictable, but may be obtained only by a direct measurement. Such a measure-
ment however disturbs the particle and thus alters its state. After the coordinate is
determined, the particle will no longer be in the state given by Eq. (2). The usual conclu-
sion from this in quantum mechanics is that when the momentum of a particle is known,
its coordinate has no physical reality.
More generally, it is shown in quantum mechanics that, if the operators correspond-
ing to two physical quantities, say A and B, do not commute, that is, if AB f. BA, then
the precise knowledge of one of them precludes such a knowledge of the other. Further-
more, any attempt to determine the latter experimentally will alter the state of the
system in such a way as to destroy the knowledge of the first.
From this follows that either (1) the quantum-mechanical description of reality
given by the wave functions is not complete or (2) when the operators corresponding to
two physical quantities do not commute the two quantities cannot have simultaneous
reality. For if both of them had simultaneous reality - and thus definite values - these
values would enter into the complete description, according to the condition of com-
pleteness. If then the wave function provided such a complete description of reality, it
60 Nathan Rosen

would contain these values; these would then be predictable. This not being the case,
we are left with the alternatives stated.
In quantum mechanics it is usually assumed that we wave function does contain a
complete description of the physical reality of the system in the state to which it cor-
responds. At first sight this assumption is entirely reasonable, for the information obtain-
able from a wave function seems to correspond exactly to what can be measured without
altering the state of the system. We shall show, however, that this assumption, together
with the criterion of reality given above, leads to a contradiction.

For this purpose the paper considers the case of two systems I and II,
which interact from the time t = 0 to t = T, after which there is no longer
any interaction between them. From the states of the systems for t < 0 one
can calculate with the help of the Schrodinger equation the wave function t/J
of the combined system I + II at a time t > T. This does not enable one to
calculate the states of the individual systems after the interaction. That can
be done only with the help of further measurements by a process known as
the reduction of the wave packet, as follows:
Let A be some physical quantity pertaining to system I with eigen-
functions Un (Xl) and corresponding eigenvalues an, where Xl stands for the
variables describing this system. (It is tacitly assumed that there is no degener-
acy.) Then t/J can be expanded in a series of the orthogonal functions Un (Xl),
00

t/J(X I ,X 2 )= L t/Jn(X 2)U n (X I ), (3)

n=l

where X2 stands for the variables describing system II and t/Jn (x 2) are merely
the coefficients of the expansion. If the quantity A is measured and is found
to have the value ak, it is concluded that I is left in the state given by the
wave function Uk (Xl) and II in the state given by t/Jk (X2)' Thus the wave
packet given by the infinite series (3) is reduced to a single term t/Jk (X2) Uk (Xl)'
This is the process of the reduction of the wave packet.
Some additional remarks are appropriate at this point. The assumption
that the measurement results in the reduction of the wave packet is some-
times referred to as the projection postulate. The validity of this assumption
depends on the nature of the measurement process. Suppose that we carry
out on a system a measurement of a physical quantity A and obtain the
value a (which must be one of its eigenvalues, according to quantum mechan-
ics). It may be that we are dealing here with a reproducible measurement,
i. e., such that an immediate repetition of the measurement is certain to give
the same result as before. In that case we can conclude that, after the first
measurement, the system is left in a state given by t/J satisfying Eq. (1) (in
order to account for the certainty of obtaining a by another measurement).
The change from the original state of the system to that given by the eigen-
function t/J represents a reduction of the wave packet. On the other hand,
one may have a measurement that is not reproducible, in which case one
Quantum-Mechanical Description of Physical Reality 61

cannot draw this conclusion. (We can think of such a measurement as one
which leaves the physical quantity with a value different from that given by
the measurement.)
One can raise the question: is it always possible to choose a repro-
ducible measurement in order to determine the value of a given observable?
For our purpose, in view of the discussion below, it is enough to consider
measurements of the position and momentum of a particle. In these cases
it seems clear that reproducible measurements are possible, in principle.
In the paper it is tacitly assumed that the measurements are repro-
ducible and hence the reductions of the wave packets take place. For the
state given by Eq. (3) this means that, if the measurement of A gives the
value ak, then system I is left in the state with the wave function Uk(XI)'
Since nothing was done to system II, the coefficient 1/!k(X2) is unaltered
(except possibly for a normalization factor arising from probability consider-
ations).
If, instead of A, we had taken another quantity B having eigenfunctions
Us (Xl) and eigenvalues bs , we should have obtained, instead of Eq. (3), the
expansIOn
00

(4)

with new coefficients ¢S(X2)' If B is now measured and found to have the
value b r , one concludes that system I is left in the state given by Ur(XI) and
II in that given by ¢r (X2)'
We see therefore that, as a consequence of two different measurements performed
upon the first system, the second system may be left in states with two different wave
functions. On the other hand, since at the time of measurement the two systems no
longer interact, no real change can take place in the second system in consequence of
anything that may be done to the first system. This is, of course, merely a statement of
what is meant by the absence of an interaction between the two systems. Thus, it is
possible to assign two different wave functions (in our example IPk and if>r) to the same
reality (the second system after the interaction with the first).

Now it may happen that the two wave functions of system II, 1/lk and
¢r, are eigenfunctions of noncommuting operators corresponding to some
physical quantities P and Q, respectively, as can be seen from the following
example:
Suppose that the two systems are two particles and that
00

(5)
- 00
62 Nathan Rosen

where x 0 is a constant. Let A be the momentum of particle I so that, accord-

ing to Eq. (2), its eigenfunctions are
up (Xl) = e(2nilh)px l, (6)

corresponding to the eigenvalue p. Since we now have a continuous spectrum,

let us write Eq. (3) in the form
00

(7)
-00

with
l/Ip(X2) = e(2nilh) (XO-X2)P. (8)
This however is the eigenfunction of
P = (bl2ni) a/ax2, (9)
corresponding to the eigenvalue - p of the momentum of particle II.
Now let B be the coordinate of particle I, with eigenfunctions
VX(XI)=O(XI-X), (10)
corresponding to the eigenvalue x, where 0 (Xl - X) is the Dirac delta-func-
tion. Then Eq. (4) becomes
00

(11)
-00

where
00

(12)
-00

This however is the eigenfunction of the operator

Q = X2, (13)

corresponding to the eigenvalue X + Xo of the coordinate of particle II. Since

PQ - QP = hl2ni, (14)

we see that it is possible for l/I k and ¢r to be eigenfunctions of two non-

commuting operators, corresponding to physical quantities.
Another example, and a very interesting one, was later given by Bohm
and Aharonov (1957). They considered the case in which the system I and
II are two particles, each having a spin ~ (in units of bI27T), while the state
of the combined system given by l/I corresponds to a total spin o. In this
Quantum-Mechanical Description of Physical Reality 63

case, if one measures the component of the spin of I in any direction, the
corresponding component of II must be equal to the measured value, but of
opposite sign. Hence, by measuring the spin of I in the x or the y direction,
one can determine the spin of II in the x or y direction. However, according
to quantum mechanics, the spins in these directions do not commute. Hence
one has a situation analogous to that in the previous example.

Returning now to the general case contemplated in Eqs. (3) and (4), we assume
that Wk and rfJr are indeed eigenfunctions of some noncommuting operators P and Q,
corresponding to the eigenvalue Pk and qr, respectively. Thus, by measuring either A
or B we are in a position to predict with certainty, and without in any way disturbing
the second system, either the value of the quantity P (that is Pk) or the value of the
quantity Q (that IS qrl. In accordance with our criterion of reality, in the first case we
must consider the quantity P as being an element of reality, in the second case the quan-
tity Q is an element of reality. But, as we have seen, both wave functions Wk and rfJr
belong to the same reality.
Previously we proved that either (1) the quantum-mechanical description of reality
given by the wave function is not complete or (2) when the operators corresponding to
two physical quantities do not commute the two quantities cannot have simultaneous
reality. Starting then with the assumption that the wave function does give a complete
description of the physical reality, we arrived at the conclusion that two physical quan-
tities, with noncommuting operators, can have simultaneous reality. Thus the negation
of (1) leads to the negation of the only other alternative (2). We are thus forced to
conclude that the quantum-mechanical description of physical reality given by wave
functions is not complete.
One could object to this conclusion on the grounds that our criterion of reality is
not sufficiently restrictive. Indeed, one would not arrive at our conclusion if one insisted
that two or more physical quantities can be regarded as simultaneous elements of reality
only when they can be simultaneously measured or predicted. On this point of view, since
either one of the other, but not both simultaneously, of the quantities P and Q can be
predicted, they are not simultaneously real. This makes the reality of P and Q depend
upon the process of measurement carried out on the first system, which does not disturb
the second system in any way. No reasonable definition of reality could be expected to
permit this.

The paper ends with the remark:

While we have thus shown that the wave function does not provide a complete
description of the physical reality, we left open the question of whether or not such a
description exists. We believe, however, that such a theory is possible.

III Bohr's Reply

The appearance of the paper was followed by a spate of replies attempt-

ing to refute its conclusions. Of all of them one will be singled out for con-
sideration here, that by Niels Bohr (1936). The justification for this choice
64 Nathan Rosen

lies in the fact that Bohr presented the "orthodox" Copenhagen interpreta-
tion of quantum mechanics, which he, more than anyone else, had developed
and which is accepted today by most of the workers in this field. The juxta-
position of the ideas of the two articles should help to understand the differ-
ences in the Weltanschauung of these two great scientists, Einstein and Bohr.
Bohr disagreed strongly with the paper and, in particular, with the
criterion of reality. According to him, the conclusion indicates the "inade-
quacy of the customary viewpoint of natural philosophy for a rational
account of physical phenomena of the type with which we are concerned in
quantum mechanics", and the criterion of reality contains an essential
ambiguity when applied to the problems under discussion.
Let us look briefly at that part of Bohr's work dealing directly with the
paper. He begins by considering the simple case of a particle passing through
a slit in a diaphragm. On the one hand, one can choose an experimental
arrangement in which the diaphragm is rigidly attached to the support so
that the slit is at a fixed position. On the other hand, one can choose an
experimental arrangement in which the diaphragm moves freely and it is
possible to determine the transfer of momentum from the particle to the
diaphragm. In the first case the slit defines the position of the particle just
after it has passed through, but there is an uncontrollable transfer of momen-
tum between particle and diaphragm so that one is ignorant of the particle's
momentum. In the second case, if one knew the particle's momentum pre-
viously, one knows its momentum after it has passed through, but its posi-
tion is unkown since one does not know the position of the slit on the (mov-
ing) diaphragm at the moment the particle passes through. The two experi-
mental arrangements, which can be thought of as suitable for predicting the
position or the momentum of the particle which has just passed through the
slit, are mutually exclusive and allow the use of complementary classical
concepts (e.g., a coordinate or a momentum) which are mutually exclusive
according to Bohr's complementarity principle.
Bohr's purpose in discussing this situation is to emphasize that one is
not dealing with an incomplete description in which one is ignorant of cer-
tain quantities, but rather that in each experimental arrangement one is
faced with the impossibility of defining certain quantities in an unambiguous
way.
Bohr then considers the problem of the two particles discussed in the
paper. According to him the situation is not very different from that of the
single particle above. In principle, one can imagine that the two particles go
through two slits in a diaphragm with an experimental arrangement for deter-
mining the momentum transferred from the particles to the diaphragm. The
distance between the slits gives x 2 - X I just after they have passed through.
Knowing the momenta of the particles before they reach the slits, one can
determine their total momentum PI + P2 after passing through. On the basis
of the commutation relations between operators corresponding to conjugate
Quantum-Mechanical Description of Physical Reality 65

variables one sees that these two operators commute, so that we can have a
state of the combined system I + II which is an eigenstate of both of them.
If we now have an additional experimental arrangement for measuring x I,
then we can determine x 2; if we have one for measuring PI, we can deter-
mine P2' This is the situation in the paper.
However, Bohr interprets this situation differently. The measurement
process used to determine x 2 prevents one from determining P2 and vice
versa. According to Bohr, the criterion of physical reality used in the paper
contains an ambiguity as regards the meaning of the expression, "without
in any way disturbing a system". In the case considered there is no mechani-
cal disturbance of system II during the last critical stage of the measuring
procedure (when one measures either XI or PI)' But there is "an influence
on the very conditions which define the possible types of predictions regard-
ing the future behavior of the system. Since these conditions constitute an
inherent element of the description of any phenomenon to which the term
'physical reality' can be properly attached, we see that the argumentation of
the mentioned authors does not justify their conclusion that quantum-
mechanical description is essentially incomplete". On the contrary, as he sees
it, this description "may be characterized as a rational utilization of all possi-
bilities of unambiguous interpretation of measurements compatible with the
finite and uncontrollable interaction between the objects and the measuring
instruments in the field of quantum theory". Therefore, according to Bohr,
the quantum-mechanical description of physical reality is complete.

IV Discussion

So what is the answer to the question: can quantum-mechanical descrip-

tion of physical reality be considered complete?
It is clear that this is not a question that can be anwered in an opera-
tional way, by means of experiments and measurements. The answer depends
on how one defines the elements of physical reality or, more generally, how
one views physical reality. It appears then that, ultimately, it is a question
of what one believes.
Einstein believed in the existence of an objective reality, independent
of the observer. With the help of measurements one can get information
about this reality, but it exists independently of these measurements (pro-
vided the measurements do not distrub or change it), and it would continue
to exist in the absence of human observes. Starting from this standpoint and
applying the criterion for an element of physical reality, the paper arrives at
the conclusion that the quantum-mechanical description is incomplete.
Bohr, on the other hand, viewed reality differently. According to him,
the elements of reality in a given system are determined by the experimental
arrangements that are set up to investigate the system. For the experimental
66 Nathan Rosen

arrangements determine the possible results of the measurements to be car-

ried out and, one might say, they therefore mold the reality into a form that
corresponds to these possible results. If one has set up an apparatus to
measure the position of a particle, this position is an element of reality while
the momentum is not, and vice versa. According to Bohr, the quantum-
mechanical description is complete because it corresponds exactly to what it
is possible to determine in a given situation, i. e., with a given experimental
arrangement. What it seems to amount to is that the description of reality
by quantum mechanics is complete because reality is whatever quantum
mechanics is capable of describing.
When Bohr considered the influence of the experimental arrangement
on the physical reality, he did not distinguish between the case in which a
direct measurement is carried out on the system of interest and the case in
which one carries out a measurement on one system (I) to get information
about another system (II). According to Einstein the two cases are quite
different; only the second case can fulfill the condition of the reality criter-
ion, "without in any way disturbing a system". To him a disturbance meant
a physical interaction with another system, not just the presence of some
measuring instrument at a distance.
Let us go back to the paper. If one accepts its conclusion (and there
are some who do), one can raise the question: what can be done to get a
complete description of reality? One possibility is to retain the present form
of quantum mechanics, which has proved to be so successful in giving agree-
ment with observation, but to supplement the information provided by the
wave function with additional information given by other quantities, often
referred to as hidden variables or hidden parameters (Belinfante 1973).
These hidden variables, if known, would help to give a complete description
of reality. For example, in the case of the two systems considered in the paper,
a knowledge of the hidden variables would tell us which term l/Ik (x z) Uk (x 1)
in Eq. (3) and which term CPr (Xz) Vr (x 1) in Eq. (4) would be obtained if A
or B were measured. However, these hidden variables are not known, and
this is what gives a statistical character to quantum mechanics.
It appears, however, that matters are not so simple. The hidden para-
meters, if they exist, give correlations between the results of measurements
carried out on system I and II after they have ceased to interact. If one
assumes the principle of "locality", i.e., that the result of a measurement
on one system does not depend on what is being done to the other system,
then it is found that the correlations determined by the hidden parameters
lead to certain statistical relations, such as Bell's inequalities (Bell 1971).
These relations can be different from those given by quantum mechanics
on the basis of correlations described by the wave function t/I in Eqs. (3)
and (4). However, experiments carried out in recent years seem to confirm
the predictions of quantum mechanics (Clauser and Shimoni 1978). It appe-
ars therefore that, if one wants to have hidden parameters, one must also
Quantum-Mechanical Description of Physical Reality 67

have some sort of non-local interaction between the systems, i. e., an inter-
action at a distance, which causes one system to be influenced by a measure-
ment carried out on the other. Most physicists would find this idea rather
unattractive.
Let us now turn to the last paragraph of the paper. What are the prospects
of finding a satisfactory theory that will give a complete description of re-
ality? One must not be overly optimistic. It appears that such a theory will
not be obtained by some simple modification of quantum mechanics, such
as the addition of hidden variables. If someday quantum mechanics is re-
placed by another theory, this is likely to involve revolutionary changes in
concepts and principles - perhaps even changes in our concepts of space and
time. In that case it may even turn out that the question posed by the paper
- is the description of physical reality complete? - no longer has a mean-
ing, or that it has to be given a different interpretation. The consequences of
a revolution in physics are hard to foresee.

References

Belinfante, F.]. 1973. A Survey of Hidden-Variables Theories, Pergamon Press, Oxford.

Bell,]. S. 1971. Foundations of Quantum Mechanics, Proc. of Int. School of Physics
"Enrico Fermi", B. d'Espagnat, ed., Course 49, p. 171, Academic Press, N.Y.
Bohm, D. and Aharonov, Y. 1957. Phys. Rev. 108, 1070.
Bohr, N. 1936. Phys. Rev. 48, 696.
Clauser,]. F. and Shimoni, A. 1978. Reports on Progress in Physics (to appear).
Einstein, A., Podolsky, B. and Rosen, N. 1935. Phys. Rev. 47, 777.
 69

Einstein's Contribution to Statistical Mechanics

Hiroshi Ezawa

Einstein's entire research career may be characterized as a search for

a unified foundation of physics!. His works on statistical mechanics were
no exceptions.
Unsatisfied with the molecular theory of James Clark Maxwell and
Ludwig Boltzmann as developed solely for gases, young Einstein worked
his way to build up 'the general molecular theory of heat'. In 1902, the
same year as he published the first of the series of three papers on the
subject, Josiah Willard Gibbs' treatise, Elementary Principles in Statistical
Mechanics, Developed with Special Reference to the Rational Foundation
of Thermodynamics, appeared. Einstein and Gibbs were working on very
much the same theme without knowing each other. Indeed, "when the time
is ripe, important ideas are developed by different men at different places,"
as Born once wrote in his enlightening account of Einstein's statistical
theories 2 •
But, there was a difference in attitudes between the two. While Gibbs
saw in the statistical mechanics the foundation of the thermodynamics as
the title of his treatise attests, Einstein sought to find in it a limit of appli-
cability of the thermodynamics, discovering as such the fluctuation phe-
nomena (1904), in particular the Brownian motion (1905) which led Jean
Perrin to achieve the first convincing proof of the reality of molecules.
Moreover, the fluctuation was made an important tool in Einstein's search
(1905-9) for implications of the Planck radiation formula (1900) upon
the nature of radiation, which led him to initiate the quantum statistical
mechanics (1907, 1924).

1. Atoms behind the Thermodynamics

As early as in 1901, when the reality of atoms was still in the dark, Ein-
stein started his research career with an attempt to determine the strengths

M. J. Klein, "Thermodynamics in Einstein's Thought", Science 157 (1967), p. 509.

2 M. Born "Einstein's statistical Theories" in Albert Einstein, Philosopher-Scientist,
Ed. by P. A. Schilpp, Evanston, Illinois (1949).
70 Hiroshi Ezawa

of interatomic forces from the available data of surface tension. His first
paper (1901) and the second (1902) were on this problem 3 ,4.
If one takes, as Einstein did already in such early days, the atomic
constitution of matter for granted, one must proceed to apply the prin-
ciples of mechanics to the atoms themselves. But, the number of atoms in
ordinary piece of matter is presumably so great that it requires statistical
methods. "At present, however, the mechanics is not yet in the position,"
says Einstein in his third paper in 1902, Kinetic Theory of Thermal Equilib-
rium and ... , "to offer sufficient foundation for the general heat theory,
for it has not yet been achieved to derive the thermal equilibrium and the
second law of thermodynamics from the equation of motion and probability
theory. "s
The method of statistical mechanics was not entirely new. Over the last
quarter of the 19 th century, Ludwig Boltzmann in Wien and James Clark
Maxwell in Edinburgh had developed the method almost fully for the case
of gases. In their kinetic theory of gases, the heat was supposed to be the
energy of molecules in chaotic flights with occasional mutual collisions,
and thermodynamic quantities of a gas were supposed to be given by the sum
or the average of corresponding mechanical quantities of the tremendous
number of molecules in the gas. Besides, the second law of thermodynamics,
which says roughly that the heat flows from the place of the higher temper-
ature to that of the lower temperature, or that the gases undergo diffusion,
was interpreted as the tendency of change into the more probable state. It
was argued by Boltzmann in fact that the change in the reverse direction
has such a small probability in general that it would never occur in the life-
time of the universe.
Einstein could not be satisfied with this state of affairs because the
theory was limited to gases only.

2. Towards the General Statistical Thermodynamics

Einstein's 1902 paper which we mentioned above was entitled Kinetic

Theory of Thermal Equilibrium and the Second Law of Thermodynam-
ics. 5 This was to be the first of a series of three papers which would
establish the statistical mechanics in such a generality that it could be used

3 A. Einstein, "Folgerungen aus den Kapillaritiitserscheinungen", Ann. d. Physik, (4),

4, p. 513 (1901).
4 A. Einstein, "Thermodynamische Theorie der Potentialdifferenz zwischen Metallen
und vollstiindig dissozierten Losungen ihrer Salze, und eine elektrische Methode zur
Erforschung der Molekularkriifte", Ann. d. Physik, (4), 8, p. 417 (1902).
5 A. Einstein, "Kinetische Theorie des Wiirmegleichgewichts und des zweiten Haupt·
satzes der Thermodynamik", Ann. d. Physik, (4),9, p. 417 (1902)
Einstein's Contribution to Statistical Mechanics 71

even beyond the realm of (classical) mechanics in which it was fostered, and
indeed it would serve as a heuristic means in the quest into the quantum
world.
That "an arbitrary physical system can be represented as a mechanical
system" was the point of departure of Einstein and indeed of the atomic
theory. It means that the state of a system is determined at any instant
of time by a point in the 2n-dimensional "phase space" spanned by the coor-
dinates q 1, ... , q n and the momenta PI' ... , Pn. The energy of the system
is a function H (p 1, ... , q n) of these variables.

Canonical and Microcanonical Ensembles. In order to apply the sta-

tistical method, Einstein considers a large ensemble consisting of N ~ 1
copies of the system each of which is in thermal contact with a huge heat
bath of a given temperature T. The state of each member of the ensemble
is represented by a point in the 2n-dimensional phase space, and hence the
state of the whole ensemble by N points. The stationary distribution should
correspond to the situation where the system is in thermal equilibrium with
the heat bath. But, what does the stationary distribution look like?
In some simple cases, the answer had been given by Maxwell and Boltz-
mann. For the monoatomic ideal gas, Maxwell showed in 1860 that their
velocities are distributed according to the exponential law in the kinetic
energy. In 1868, Boltzmann 6 generalized this law into the law for complex
molecules in the presence of an external force by elaborate calculations.
In his 1902 paper, Einstein derived the general result, the exponential
law in the total energy,
dN = ."
,A e-~H (PI," ' qn) dp 1 ... d qn· (1)

in a simple fashion characteristic to him; where ~ = lIkT and the constant

A is to be determined such that the total number of the members of the en-
semble is N Today, we call the ensemble having the distribution (1) over
the phase space the canonical ensemble after Gibbs, who arrived at the
same result independently. The constant k is called by the name of Boltz-
mann, though he never wrote it out.
Einstein begins his derivation of (1) by considering a system in contact
with the heat bath, assuming that the energy of the combined system is the
sum of the energies, thereby neglecting the supposedly small interaction
energy. He then looks at the time development of the system. But, since
Liouvilles theorem, valid for any Hamiltonian system, says that the phase
volume does not change with time, density of the representative points is
conserved. This much had been known to Boltzmann already.

6 L. Boltzmann, Wien. Ber. 63, p. 397 (1871). See also "Vorlesungen iiber Gastheorie",
J. A. Barth, Leipzig, 1895-98, Teill-lII.
72 Hiroshi Ezawa

Now, Einstein brings out a basic assumption which is made even in the
statistical mechanics of today: "We assume that, except for the energy or its
functions, there are no such functions of the state variables p, q, ... only, that
do not change with time." Then, it follows that the density distribution
of the representative points in the phase space must be a function of the to-
tal energy only. If all the system in the ensemble have the same energy, then
the density distribution of the points in the phase space is uniform over the
energy shell. Boltzmann in his gas theory stated this proposition as ergodic
hypothesis (1884,1887).
Now, the distribution of the subsystems in their phase space is obtained
if one integrates the joint distribution over all the possible phases of the
other subsystems, that is, the heat bath. Einstein finally arrives at a constant,
~, that determines so-to-say the equilibrium of the system and the heat
bath. The value of ~ is then shown to have the properties of the temperature:
First, it is always positive. Second, if there are different systems in thermal
equilibrium with one and the same heat bath, they share the same value of ~.
Taking, then, the ideal gas as one of those systems, 1/~ can be identified with
2/3 times the average of the kinetic energy of a gas molecule, which is
known from the equipartition theorem to equal 3/2 RT divided by the num-
ber NA of molecules in a mole. Thus, one gets ~ = l/kT with k = RINA .
Actually, this last determination of k was done by Einstein in his 1904 paper
to which we shall come later.
In contrast to the canonical ensemble of systems in thermal equilibrium
with a heat bath, the ensemble of isolated systems having a common energy
is called microcanonical. Remarkably, Einstein's treatment shows that the
two descriptions, microcanonical and canonical, are equivalent statistically,
if one is interested in a physical quantity of a relatively small subsystem.

Beyond the Mechanics. In the next paper of 1903, A Theory of Foun-

dation of the Thermodynamics 7 , Einstein showed that the following general
assumptions are enough to prove the microcanonical and canonical distribu-
tions in the phase space of appropriately chosen state variables. The assump-
tions are; (1) The present state determines the future (causality) by some
differential equations, (2) a system approaches a stationary state if left
isolated for a sufficiently long time and (3) the energy is the only conserved
quantity.
This generalization must have been crucial when Einstein proceeded
to apply his statistical thermodynamics to radiation without as yet having
the Hamiltonian formalism of radiation at hand.

7 A. Einstein, "Theorie der Grundlagen der Thermodynamik", Ann. d. Physik, (4), Ii,
p.170(1903).
Einstein's Contribution to Statistical Mechanics 73

Entropy and the Second Law. There is another important aspect of

Einstein's statistical thermodynamics which we have left untouched so far.
It is the mechanical definition of the entropy.
Thermodynamically, we say that the entropy of a system increases by
the amount dQIT when the heat dQ is transferred to it from a heat bath of
temperature T almost in thermal equilibrium with the system. In his 1902
paper, the first of the series on statistical thermodynamics, he treated the
heat transfer mechanically, assuming that it was mediated by the forces the
molecules of the heat bath exert upon the molecules of the system. But,
such forces can include, e.g., the pressure that does not transfer heat.
Einstein distinguished two kinds of external forces upon the system:
"The forces of the first kind are those which represent the conditions on the
system, and are derivable from a potential which is a function of the coor-
dinates q 1, ... , qn only (adiabatic walls, gravitational force etc.)." Then,
"The forces of the second kind are not derivable from a potential which
depends only on the coordinates q 1, ... , qn'" He had added, in the intro-
duction of this paper, that these are rapidly changing.
Identifying the heat transfer dQ with the work done by the latter
forces and relating the temperature T to the average kinetic energy of the
molecules of the system by the equi-partition theorem, he could show
that dQIT was a total differential in accordance with the second law of
thermodynamics. Einstein thus arrived at the entropy of the system in
thermal equilibrium, and concluded, "The second law comes out as a neces-
sary consequence of the mechanical world view."
However, it was only in the second paper (1903) of the series that he
proved the other half of the second law which says that the entropy of an
isolated system is ever increasing (or, more precisely, non-decreasing).
To prove this entropy theorem, he related the entropy to probability
by looking in two ways at an isolated system L which consisted of many
subsystems, 01, O 2 , ... , thermally insulated from each other by adiabatic
walls. It is divided into small parts Or such that the temperature in each Or
be regarded uniform.
First, he looks at the system L as a whole. He considers its phase space
and divide the energy shell of energy E into L cells of equal volumes. Given
N ~ 1 copies of L, all having the same energy E, the probability to find,
n 1 of them in the first cell, n 2 in the second, ... is (1/ L)N times W with
(2)

because, for the microcanonical ensemble, the equal volumes on the energy
shell are equally probable. Boltzmann defined a number similar to Wand
called it the number of Komplexions 8 . In general, the number of Komple-

8 L. Boltzmann, "Uber die Beziehung zwischen dem zweiten Hauptsatz der mechani-
schen Wiirmetheorie und der Wahrscheinlichkeitsrechnung respektive den Siitzen iiber
Wiirmegleichgewicht". Wien. Ber. 76, p. 373 (1977).
74 Hiroshi Ezawa

xions may be defined as the number of the microscopic states (Komplexions)

that belong to one and the same macroscopic state.
Second, he looks at the rth subsystem ar of each copy of ~, assuming
that it is in a stationary state. He adds: "The distribution of the states of the
system ar will not be noticeably different from the distribution which would
arise if the system ar were in thermal contact with a physical system of the
same temperature." This remark permits one to obtain the entropy S of the
system and numbers nr in (2). It turns out that the number of Komplexions
(2) is related to the entropy of the ensemble by

S = k log(WIL N ), (3)
where an additive constant is immaterial for the (classical) entropy. These
relations between the entropy and the number of Komplexions or the prob-
ability were first discovered by Boltzmann in the gas theory context (1877),
and are called the Boltzmann principle.
Now, Einstein says: "We have to assume that the more probable distri-
bution of the states will always follow the less probable one." From this
assumption, there follows the desired entropy theorem, namely the increase
in the entropy. Einstein gave another proof in the third paper (1904) of his
series 9 , this time using (1) as the bridge to the probability.
In 1911, Paul Hertz criticized the proof sayinglo: "If one assumes, as
Einstein did, that the more probable distribution follows the less probable,
he is introducing an assumption which has no evidence and needs to be pro-
ven." We all know that here is the core of the problem of the second law
of thermodynamics. "I take this criticism to be completely right"ll replied
Einstein in Remarks on P. Hertz's Paper (1911). As a matter of fact, he had
already realized that the changes in the reverse direction, the less probable
following the more probable, could take place in the fluctuation phenome-
na.

3. Fluctuation Formula

When a physical system 1S m thermal equilibrium with a huge heat

bath, their interaction, however weak it may be, causes exchange of energy
between them. Consequently, the energy content of the system fluctuates.
Some other variables may fluctuate also. If there were a system that had not

9 A. Einstein, "Zur allgemeinen molekularen Theorie der Wiirme", Ann. d. Physik, (4),
14, p. 354 (1904).
10 P. Hertz, "Ober die mechanischen Grundlagen der Thermodynamik", Ann. d. Physik
33, p. 225 u. S. 537 (1910).
11 A. Einstein, "Bemerkungen zu den P. Hertz'schen Arbeiten: Mechanische Grundlagen
der Thermodynamik", Ann. d. Physik, 34, p. 175 (1911).
Einstein's Contribution to Statistical Mechanics 75

too small fluctuations in one or other of its thermodynamical variables, it

would mean a limit of applicability of thermodynamics which treats the
average values only.

Fluctuation of Energy. In his paper, On the General Molecular Theory

of Heat, (1904), Einstein derived a formula for energy fluctuation of a sy-
stem in thermal equilibrium with a heat bath. On the basis of the exponen-
tiallaw (1) Einstein gets the formula *

(4)

The validity of this formula is not restricted to mechanical systems.

Thus, Einstein sees the general significance of the constant k here, i.e.,
in that it determines the magnitude of fluctuation, and hence the thermal
stability of a system.
The energy fluctuation in ordinary piece of matter is too small to be
observed. To Boltzmann and Gibbs, indeed, the small fluctuation was the
basis of their statistical approaches.

Application to Radiation. However, Einstein points out, "After all,

we may suppose that the energy fluctuation is significant only in one kind
of physical systems within our experiences." It is a cavity filled with the
thermal radiation. In fact, he could show that the Wien displacement law 12 ,
can be deduced, though approximately, from the Stefan-Boltzmann law,
and the natural assumption that the fluctuation be the greatest, if the wave
length at the spectral peak is comparable with the size of the cavity. This
success must have strengthened Einstein's conviction in the generality and
the power of his statistical thermodynamics.

4. Theory of Brownian Motion

There is a kind of fluctuation that is visible! "It will be shown in this

paper" wrote Einstein in 1905, "that ... , as a result of the molecular motion
of heat, a particle of microscopic size must perform a motion of such a size
that the motion can easily be detected with a microscope."13 This is the

The bar denotes the average value of the quantity considered.

12 W. Wien, "Temperatur und Entropie der Strahlung", Wied. Ann. d. Phys. 52, p. 132
(1894 )
13 A. Einstein, "Die von der molekularkinetischen Theorie der Warme geforderte Bewe-
gung von in ruhenden Fliissigkeiten suspendierten Teilchen", Ann. d. Physik, (4),17,
p. 549 (1905)
76 Hiroshi Ezawa

Brownian motion, though at the time of writing the paper Einstein could
not be sure about the identification because "the informations available
to me ... was so inaccurate."
Einstein asserted, this motion, if ever observed, would tell us two
things. One is that "the classical thermodynamics cannot be taken as exactly
valid already in microscopically discernible space." The other is that "an
exact determination of the true sizes of atoms becomes possible."

Brownian Motion. It is not clear how much was known to Einstein

about the Brownian motion. Back in 1828, Robert Brown, an English
botanist, observed under his simple microscope that those minute particles
which were ejected into water by a water-swollen pollen exhibited a vigor-
ous trembling motion. He discovered soon that the trembling motion was
a universal phenomena, exhibited by any fine particles, and the universal
motion became to be called the Brownian motion. In the last quarter of the
19 th century, some atomists suspected that the Brownian motion was caused
by the collisions of water molecules in random thermal motion.
In order to check the molecular theory of the Brownian motion, direct
measurement of particle velocities were attempted with negative results that
they were much smaller than the values expected from the equi-partition
theorem of energy. Moreover, it was subsequently found out that the appar-
ent mean velocities of the Brownian particles varied in magnitudes depend-
ing upon the length of the time t of observation, without approaching a limit
as t....,. o.
Einstein had/his own point of view. In his famous paper!3 of 1905, he
shows that for the mean square displacement (.6X)2 of a particle (radius a)
suspended in a fluid medium (temperature T and viscosity 'T}) the following
simple relation holds:

(LlX)2 = 2Dt, D = kT/67r'T}a. (5)

Because Einstein's (5) relates the fluctuation (.6X)2 with the coefficient 17 of
friction, or energy dissipation, it is a prototype of the fundamental theorem
- the fluctuation-dissipation theorem - of today's statistical mechanics of
irreversible processes 14 .
Time was ripe for looking into the molecular constitution of matter.
As soon as Einstein's 1905 paper was published, endeavours started to link
his theory to experimental observations.
However, the theory was not very well understood. The first point
many people wanted to check experimentally was the equipartition theorem

14 R. Kubo, "Fluctuation-Dissipation Theorem", Rept. on Prog. in Phys. 1, p. 255

(1966).
Einstein's Contribution to Statistical Mechanics 77

Einstein applied to the Brownian particles. In view of Svedberg'slS unsuccess-

ful attempt (1906), in particular, Einstein had to write Remarks on the
Brownian Motion 16 (1907) to point out that the velocity of the Brownian
particle would change in magnitude and direction all too often to permit any
measurements.
Jean Perrin, who had been probing the molecular theory in regard to
the height distribution of colloidal particles, turned to examine the displace-
ment formula (5) experimentally. His results were consistent with the theory,
and determines the value of the diffusion constant D. The radius of the par-
ticle was determined carefully by Perrin's ingeneous method, and the Ein-
stein relation (5) could be used to find out the value of the Boltzmann con-
stant k and then the value of the Avogadro number NA . Upon receipt of
Perrin's paper, The Brownian Motion and the Molecular Constants 17 , Ein-
stein wrote back 18 :
"I had thought it impossible to investigate Brownian movement so pre-
cisely; it is fortunate for this matter that you have taken it upon yourself ....
A precise determination of the size of the molecule appears to me of
the highest importance moreover, because the radiation formula of Planck
can be more sharply proved through this than through measurements of ra-
diation. The Planck theory of radiation also leads to a determination of the
absolute size of atoms, with exact validity claimed."
Establishing the reality of atoms marked a turning point of physics,
which one can see most clearly in the lectures at the 1st Solvay Conference
in 1911. Perrin lectured on 'The Proof of Molecular Reality', but most of
the other lectures dealt with the 'quanta'. The title of Einstein's lecture
was 'The Present Situation of the Problems of Specific Heat.' The specific
heat shall be our subject of the next section.

5. If Light is Quantized, then Mechanical Oscillations Also

In Einstein's paper in 1907, The Planck Theory of the Radiation and

the Theory of Specific Heat 19 , we find him declare, "While we have so far

15 Svedberg, "Ober die Eigenbewegung der Teilchen in kolloidalen Liisungen", Zeit-

schr. f. Elektrochemie, 12, p. 853 (1906).
16 A. Einstein, "Theoretische Bemerkungen tiber die Brownsche Bewegung", Zeitschr.
f. Elektrochemie, 13, p. 41 (1907).
17 J. Perrin, "Mouvement Brownien et constantes moleculaires", Comptes Rendus 149,
p. 477 (1909).
18 Unpublished letter quoted in the book by M. Jo Nye, "Molecular, Reality, a Perspec-
tive on the Scientific Work of Jean Perrin", MacDonald, London and American
Elsevier, New York (1972), p. 135.
19 A. Einstein, "Die Planck'sche Theorie der Strahlung und die Theorie der spezifischen
Wiirme", Ann. d. Physik, (4), 22, p. 180 and p. 800 (correction) (1907).
78 Hiroshi Ezawa

regarded the molecular motion to obey the same laws as those valid for the
perceptible bodies, it is now necessary to assume for the oscillating ions
that the manifold of states that they can take is smaller than that for the
bodies within our experiences." This was the beginning of the quantum
statistical mechanics. Einstein was led to this conclusion by the following
path of thoughts.
It was from the end of the 19 th century that the spectrum of black
body radiation had been a central issue of physicists. In 1900, Planck 20
discovered his radiation formula which, with an appropriate choice of the
contained parameter h, now called Planck's constant, fits very well with
the measured energy density of the black body radiation. He could even
derive the formula on the assumption that the radiation of frequency v
was emitted or absorbed in a lump, or energy quantum of hv, although
his continuous efforts to understand the mechanism of the discontinuous
changes were never successful.
Einstein could not accept Planck's derivation of his formula. Even
as late as in 1911, after Planck's report in the Solvay Conference, Einstein
expressed his opinion that Planck's probability calculation was a shock to
him. Let us see how Planck derived his formula in his 1900 paper.

Planck's Komplexions. Planck considered a set of resonators, i.e.,

charged harmonic oscillators, in thermal equilibrium with the radiation
field. The energy density of the radiation field is calculated from the average
energy E (v) of the resonator from the electromagnetic theory.
The cavity has many resonators of different frequencies Vs. By the
assumption of the discontinuous energy changes, the energy of each resona-
tor can be taken to be an integer multiple of hv s . Let Ns and Ps . hv s be the
number and the total energy of the resonators of frequencies Vs' Then, in
how many different ways can the energy Ps . hvs be distributed over the Ns
resonators, S = 0, 1,2, ... ? Planck's answer for the number of the Komple-
XlOns 1S
w =1T (Ns + Ps - 1)!
(6)
N s !(Ps -l)!
s
The state of thermal equilibrium of the resonator system is the state of the
largest 'probability' W, or equivalently the state with the largest entropy S
under the condition that the total energy E is given. This determines Ps and
hence the average resonator energy E(v s ) = PshvJNs .
To Einstein, however, it was not clear how Planck could justify his
assumption that his Komplexions are equally probable to each other.

20 M. Planck, "Ober eine Verbesserung der Wien'schen Spektralgleichung", Verh. dtsch.

phys. Gesellschaft 2, p. 202 (1900).
Einstein's Contribution to Statistical Mechanics 79

Light Quanta Faced with the above difficulties, he proposed in On a

Heuristic Viewpoint regarding the Generation and Transformation of Light
(1905)21, to consider the problem "keeping a close contact with experiences
and without employing as a basis any picture of the generation and propaga-
tion of the radiation."
The Rayleigh-Jeans formula disagrees with experiments, but not in the
whole range of the spectrum. It is good in the low frequency limit, and in
this limit v -+ 0 in fact the correct Planck formula reduces to Rayleigh-Jeans'.
Einstein looked at the opposite limit v -+ 00, where something novel should
be. In this limit, the Planck formula reduces to its predecessor, Wien's
formula (1896), which he found to imply that the radiation of high fre-
quency v behaves like a collection of particles of energy hv. Indeed, by the
entropy calculation based upon Wien's formula, the probability that a mono-
chromatic radiation of frequency v, energy E and in a large cavity of volume
V happens to be completely localized in a small portion of volume v was
found to be
V )ElhV
Wradiation = (V ' (7)

which would be the same as the corresponding probability for the ideal gas
if E/hv were the number of gas particles.
Einstein saw supporting evidences for this particle analogy of radiation
in the Stokes law of photoluminescence, the photoelectric effect and the
ionization of gases. Thus, he suggested as a heuristic point of view the "As-
sumption that the energy of light is distributed discontinuously in the space .
... When the light propagates from a point, its energy will not be continuous-
ly distributed over an ever increasing volume of space, instead it consists
invariably of a finite number of energy quanta which can move without
disintegrating and can be absorbed or generated as a whole." This is the light
quantum hypothesis.
Such a view might seem refutable at once by the ample evidences of the
wave theory of light. So Einstein remarked "that the optical observations are
related to the values averaged in time, and not to the values at a moment",
and therefore "it is ... conceivable that the theory employing continuous
functions over a space leads to contradictions with experiences, if one
applies it to the phenomena of generation and transformation of light."

Assumptions Underlying the Planck Theory. In the subsequent paper,

On the Theory of Generation and Absorption of Light (1906)22, he turned

21 A. Einstein, "Ober einen die Erzeugung und Verwandlung des Lichts betreffenden
heuristischen Gesichtspunkt", Ann. d. Physik, (4), 17, p. 132 (1905).
22 A. Einstein, "Zur Theorie der Lichterzeugung und Lichtabsorption", Ann. d. Physik,
(4),20, p. 199 (1906).
80 Hiroshi Ezawa

to examine the entropy of the resonators, finding that his formula (1) gives
the entropy in accord with Planck's number of Komplexions if:

[ 1] The energy of an elementary resonator (of frequency v) can take only

the values which are integer multiples of hv.
This assumption implies that "the energy of a resonator changes discontinu-
ously by absorption and emission." and hence that Maxwell's electrody-
namics is not applicable to these processes. Nevertheless, one has to assume:
[2] The average energy of a resonator immersed in radiation is equal to
the value that one calculates using Maxwell's theory of electricity.
Recall that the entropy of the radiation required also
[3] The light quantum hypothesis.

Einstein has thus found three assumptions that underlies Planck's theory of
radiation. He showed that Planck's formula follows from these assumptions
in an impressively simple way.
Logically, one may question if both assumptions [ 1] and [3] are needed.
To Einstein, who saw "disturbing dualism" in particle and field concepts in
classical physics 23 , it was probably most natural to assume that, on the fun-
damental level, what is true with radiation should also be true with mechani-
cal objects and vice versa.

Quanta in the Purely Mechanical Regime. It was in this spirit of unify-

ing physics that Einstein proceeded to apply the assumption [1] above to
the thermal vibration of atoms in solids. This was by no means a small step,
as T. S. Kuhn 24 shows by his extensive historical studies. People, including
Planck himself, recognized the need for the quantum h in discretizing the
energy or the phase space for the purpose of calculating probabilities. But,
to Planck, say, explaining that need was a challenge to be resolved by the
further development of the theory of electron-radiation interaction. It
was looked at with caution that Einstein extended the quantum concept
to the atomic vibrations which have nothing to do with the radiation.
In his 1907 paper which we have referred to at the beginning of this
section, Einstein noticed, "If the Planck theory of radiation hits the heart
of the matter, then we must expect to find, also in some other domains

23 A. Einstein, "Autobiographical Notes" in Albert Einstein, Philosopher-Scientist,

Ed. by P. A. Schilpp, Evanston, Illinois, 1949.
24 T. S. Kuhn, "The Quantum Theory of Specific Heats: A Problem in Professional
Recognition", Proc. of the XIV. Congress of the History of Science, No.1 (1974),
No.4 (1975), Science Council of Japan, also M.J. Klein, "Einstein, Specific Heat
and the Early Quantum Theory", Science 148, p. 173-180 (1965).
Einstein's Contribution to Statistical Mechanics 81

of heat theory, contradictions between the present-day molecular kinetic

theory and the experiences. "21
Indeed, in the statistical mechanical theory of specific heat, there
were difficulties related to the equi-partition theorem, as Einstein pointed
them out. In the simplest model of the thermal motion in solids, i. e., the
simple harmonic motions of atoms around their own equilibrium positions,
the equipartition theorem gives c = 3 R for the specific heat per mole. Many
of the substances in solid state obey this rule (Dulong-Petit's law, 1819).
However

(a) The solids of carbon, boron and silicon have specific heats significantly
smaller than 3 R.
(b) There is no indication of the added contribution to the specific heat
from the electrons, which according to Drude's analysis of optical
dispersion should exist in solids and should be capable of oscillations
with ultraviolet eigenfrequencies.

Now, by Einstein's generalization of the assumption [1] to the oscilla-

tors in solids, shows that the specific heat of such oscillators is a function
of kTlhv. Remarkably, the specific heat decreases to zero for low tempera-
ture, while at high temperature the Dulong-Petit law is recovered. Thus, it
was immediately clear to him that the difficulty (b) was resolved; For the
ultraviolet eigenfrequencies of electrons, one has kTlhv <{ 1 at room tem-
perature, so that there is no electron contribution to the specific heat.
(This is ascribed today to the Fermi-Dirac statistics of electrons.) As for
the difficulty (a), all he could was to point out that those elements in
question all had small atomic weights and rather high observed infrared
eigenfrequencies.
As a matter of fact, experimental evidences for the Einstein theory of
specific heat remained extremely scarce until 1910 and even physicists
favorable to the Planck theory of radiation ignored Einstein's generaliza-
tion.2s
We find the impact of the Einstein theory in the letter Walter Nernst
wrote to Ernest Solvay on July 26, 1910. Nernst was a leading physical
chemist in Berlin, working at that time to organize the 1st Solvay Confer-
ence. He wrote "Wen are currently in the midst of a revolutionary refor-
mulation of the foundation of the kinetic theory of matter.' '26 The measure-
ments of his own group as prompted by his discovery of the third law of
thermodynamics had shown that the specific heats of solids, almost with-
out exception, had temperature dependence in contradiction to the equi-

25 T. S. Kuhn, ibid.
26 quoted in detail by T. S. Kuhn, ibid.
82 Hiroshi Ezawa

partition theorem, tending to zero as T ..... 0 in accordance with the third

law. These contradictions were resolved if one restricted the motion of
electrons and atoms by the 'doctorine of energy quanta', Thus, Einstein's
generalization of the assumption [1] to purely mechanical regime was
accepted, opening up a new era of quantum theory. It was in fact here
that people began to feel the need for formulating a more general quantum
condition which is applicable not only to harmonic oscillators but more
generally to any mechanical systems.

6. BQse-Einstein Statistics and Particle-Wave Duality

In his paper in 1909, On tbe Present Situation of tbe Problems of

Radiation 27 , Einstein was still asking "how is the Planck theory of radia-
tion related to the theory which rests upon the theoretical foundation as
recognized today?"
In particular, he could not be satisfied with the derivations, neither
Planck's nor his own, of the radiation formula. These derivations applied
the statistical mechanics to the resonators, the result being then transferred
to the side of the radiation by the assumption [2] of classical nature, where-
as, from the point of view of his statistical mechanics, it must have been
possible to treat the radiation directly. Besides, Planck's use of his Kom-
plexion was questionable in two respects: First, as mentioned before, there
was no guarantee that each of the Komplexion Planck considered was
equally probable. Second, it appeared meaningless to talk about distri-
buting the energy quanta bv over different resonators, because by Planck's
result itself bv was much larger than the average energy E (v) of the reso-
nators.

Particle-Wave Duality. Despite its unsatisfactory derivation, the Planck

formula was firmly confirmed by experiments. So, Einstein kept trying to
deduce its implications upon the nature of radiation. He reconsidered that
problem of fluctuation of the radiation in a small portion v of the cavity,
which he had once treated using Wien's formula. The entropy consideration
based this time upon the full Planck formula yielded a result which demon-
strated the duality of radiation - its particle nature as well as its wave nature.
He found it remarkable "that the characteristics of both structures (wave
structure and quantum structure) ... must be looked upon not as mutually

27 A. Einstein, "Zum gegenwartigen Stand des Strahlungsproblems", Physikalische Zeit-

schrift 10, p. 185 (1909).
Einstein's Contribution to Statistical Mechanics 83

incompatible. "28 This finding, however, did not help him to revise the deri-
vation of the Planck formula.
Einstein's interests gradually shifted to relativity and gravitation.

Transition Probability. In the meantime, Niels Bohr29 put forward his

theory of atomic structure (1913), in which discrete energy levels were
singled out from the continuous manifold of classical orbits by the Bohr
'quantum condition' and the radiation were assumed to be emitted or ab-
sorbed by quantum transitions of electrons from one of atom's discrete
energy levels to another. Based on this idea, Einstein showed in 1916 and
1917 that the Planck formula could be derived by using a rate equation 30 , 31.
Here, he introduced the concept of transition probability and the concept
of spontaneous emission. Yet, this derivation was not in line with his sta-
tistical mechanics.

Mutual Influence of Puzzling Nature. In 1924, a paper entitled Planck's

Law and the Light Quantum Hypothesis 32 , appeared, giving the derivation
of the Planck formula that applied the statistics to the radiation directly.
It was the paper by an Indian physicist, Satyendranath Bose, which Einstein
translated into German to send for publication on the request of the author.
Following Planck, Bose divided the whole phase space into small cells of
volume h 3 , assuming that only one state of a photon (of given polarization)
belongs to each cell.
Now, an instantaneous state of the black body radiation, i.e., a Kom-
plexion, is represented by a distribution of many representative points over
the phase space. Let p~ be the number of those cells which lie in the energy
shell corresponding to the frequency interval (v s , Vs + dv) and which have k
representative points each. Then, how many different Komplexions belong
to such a distribution? Or, in other words, in how many different ways can
one realize such a distribution? Bose's answer:

wBose =TI s, gs!s, (8)

s PO' PI .
where p~ + p~ + ... = gs is twice the total number of the cells in the energy
shell; 'twice' because two independent polarizations of a light quantum are

28 A. Einstein, "Ober die Entwicklung unserer Anschauung iiber das Wesen und die
Konstitution der Strahlung", Physikalische Zeitschrift, 10, p. 817 (1909).
29 N. Bohr, "On the Constitution of Atoms and Molecules", Phil. Mag.21, p.l (1913).
30 A. Einstein, "Strahlungsemission und -absorption nach der Quantentheorie" Deutsche
Physikalische Gesellschaft, Verhandlungen, 18, p. 318 (1916).
31 A. Einstein, "Quantentheorie der Strahlung", Physikalische Zeitschrift, 18, p. 121.
32 S. Bosse, "Planck's Gesetz und Lichtquantenhypothese", Zeitschrift fiir Physik 26,
p. 178 (1924).
84 Hiroshi Ezawa

possible in each cell. Bose showed that the Planck formula corresponds to
the 'most probable' distribution.
Many people were perplexed by the counting (8), Einstein himself,
Ehrenfest and Schrodinger among them.
Ehrenfest and others were concerned that in Bose's theory the quanta
were not treated as statistically independent from each other. But, Einstein
could show that Bose's formula (8) was equivalent to Planck's (6). This
implied that Planck's assumption of equal probability for his Komplexions
was not compatible, as Einstein had long suspected, with the assumption
of statistical independence of light quanta; there must be "a mutual in-
fluence ... which, for the moment, is completely puzzling."

Bose-Einstein Statistics. It was after the advent of the quantum mecha-

nics, in particular the many-body theory that the real significance of the
'mutual influence' was found to lie in the indistinguishability of like par-
ticles. Einstein found that Bose's formula (8), when applied to the ideal
gas, resolved two problems of Boltzmann's namely the Gibbs' paradox
and the conflict with the experimentally established third law of thermo-
dynamics. To check the latter, we have only to notice that at T = 0 all
the gas molecules sit in the cell of the lowest energy (s = 0, say) which is
unique, go = 1, so that WBose = 1. Note that W Boll1lnann = N! due to the
distinguishability presumption. Thus, in the second of his series of two
papers, Quantum Theory of a Monoatomic Gas. 33 he declared that, in
place of Boltzmann's formula, Bose's should be used for gases as well as
for radiation. The "deep kinship between radiation and gas" thus disco-
vered was indeed in accord with his spirit of unifying physics.
Bose's way of counting the number of Komplexions so generalized
is today called the Bose-Einstein statistics. It can be characterized, as we
mentioned above, by saying that the gas particles are indistinguishable and
any 'quantum state' (or the cell) can be occupied by any number of partic-
les. When 'any number of' is replaced by 'at most one', we get the Fermi-
Dirac statistics which was formulated by Enrico Fermi in his study of
Nernst's third law of thermodynamics 34 .

Dawn of the Wave Mechanics. We recall that the fluctuation formula

of the radiation energy manifested the particle-wave duality of radiation.
But, it was obtained from formula (6), which is equivalent to (8). Therefore,
the particle-wave duality is implied by the Bose-Einstein statistics, so that,
as Einstein pointed out, it must apply to the gas molecules, too.

33 A. Einstein, "Quantentheorie des idealen Gases", Preussische Akademie der Wissen-

schaft, Phys.-math. Klasse, Sitzungsberichte, 1925, p. 18.
34 E. Fermi, "Zur Quantelung des idealen einatomigen Gases", ZS. f. Physik, 36, p. 902
(1926).
Einstein's Contribution to Statistical Mechanics 85

That a material particle has a wave aspect was in accord with the hypo-
thesis Louis de Broglie 35 put forward in the thesis he defended on November
29, 1924, just before Einstein completed the part II of his paper (received
in December 1924). In Einstein's words in his 1925 paper: "Mr. L. de Broglie
has shown in his very remarkable paper how a (scalar) wave field can be
associated with a material particle or a system of material particles." Further
in the footnote attached to this sentence: "In this thesis, there is also a very
remarkable geometrical interpretation of the Bohr-Sommerfeld's quantiza-
tion rule. "32 Note the 'very remarkable' used twice. Indeed, it was Einstein
who called attention of many people to de Broglie's work; the paper itself
was published with his encouragement. 36
Schrodinger pointed out at the end of 1925 that a deeper understand-
ing of Einstein's new theory would be obtained if one associated each energy
state Es of a single molecule with a degree of freedom of the whole system
and said that the 5 th degree of freedom had, like a harmonic oscillator, the
energy ns Es in place of saying that ns molecules were in the state Es. His
point was that in this picture, the indistinguishability of Bose-Einstein
'particles' becomes obvious within the ordinary, well established statistics.
"The real meaning of Einstein's gas theory is", wrote Schrodinger 37 , "that
the gas is to be understood as such a system with linear eigen-oscillations as
a volume of radiation or a solid body." Not more than one month later, the
1't communication of Schrodinger's series of papers38, 'Quantization as an
Eigenvalue Problem', was received for publication in Annalen der Physik.
Einstein's influence went so far as to help conceive the wave mechanics.

Bose-Einstein Condensation. Another important prediction from

Einstein's theory of ideal gas is that below a critical temperature some of
the gas molecules condense to the state of zero momentum. People did not
take this prediction seriously because it appeared that any gas would liqui-
dify at such low temperatures.
In 1938, fourteen years after Einstein's paper, Fritz London 39 aroused
interests in the Bose-Einstein condensation as he called it, by relating it to
the peculiar behaviours of liquid helium, e.g., the superfluidity. London
could account qualitatively for the peculiar behaviours of the liquid helium

35 L. de Broglie, "Recherches sur la Theorie des Quantes", Theses, Paris 1924 und Ann.
d. Physik (10),3, p. 22 (1925).
36 W. Heitier, "Erwin Schrodinger", Biographical Memoirs of Fellows of the Royal
Society 1961 (Va. 7), The Royal Society, Burlington House, London.
37 E. Schrodinger, "Zur Einstein'schen Gastheorie", Phys. Zeitschrift, 27, p.101 (1926).
38 E. Schrodinger, "Quantisierung als Eigenwertproblem", Ann. d. Physik (4), 79,
p. 361,p.489;80,p.437;81,p. 109(1926).
39 F. London, "On the Bose-Einstein Condensation", Phys. Rev., 54, p. 947 (1938).
86 Hiroshi Ezawa

- superfluidity, reversible heat conduction and fountain effect - by taking

a 'two-fluid picture', one fluid consisting of the condensates and the other
of thermal molecules.
Although London's theory had to be substantially modified by taking
into account the interatomic forces, the relevance of the Bose-Einstein con-
densation remained. Today, the concept of the Bose-Einstein condensation
has a very important place in statistical physics as well as in elementary
particle physics. It has been found that not only the Bose-Einstein particles
condense but also do the Fermi-Dirac particles by forming a pair called the
Cooper pair, after Leon N. Cooper who noticed the possibility in his efforts
to understand the superconductivity.

7. Concluding Remarks

What were Einstein's contributions to statistical mechanics? Before

Einstein, the statistical methods were almost there in the kinetic theory
of gases by Maxwell and Boltzmann, and the detailed analysis of who went
how far has to be left to historians of science.
Nevertheless, it may be said that, unlike Boltzmann who struggled to
grasp the irreversible time development of a state of a gas, Einstein as well
as Gibbs focused their attention to the stationary states thereby isolating
successfully the methods of equilibrium statistical mechanics as embodied
in their systematic use of the canonical ensembles. Due to this restriction,
however, they could not treat the problem of approach to equilibrium, or
the second law of thermodynamics, dynamically. Einstein attempted to
deduce the second law, but only on the assumption that states change
towards the more probable ones. The same idea can be found in Boltz-
mann (1879).
Referring to Gibbs' treatise, Elementary Principles in Statistical Me-
chanics (1902), Einstein wrote in 1911, "If Gibbs' book had been known
to me at that time, I would not have published those works and instead
restricted myself on a few problems." 11 But, it was fortunate that he could
work his way40, because the develqpment of the quantum theory that follow-
ed owed very much to the intuition and tools he developed in his endeavor
to construct his 'general molecular theory of heat'. The following are the
contributions that are distinctively Einstein's:
First, Einstein was able to point to a case of observable fluctuation,
namely the Brownian motion, which led to the first convincing proof of the

40 Gibbs' book was translated into German in 1905: J. w. Gibbs, "Elementare Grund-
lagen der statistischen Mechanik", translated by E. Zermelo, J. A. Barth, Leipzig
1905.
Einstein's Contribution to Statistical Mechanics 87

reality of molecules. Further, he made the fluctuation a useful theoretical

tool in his quest into the fundamental nature of radiation and matter.
Second, Einstein generalized the framework of his statistical thermo-
dynamics beyond the range of mechanics. The generality of the framework
must have given him confidence when he applied it to the radiation for
which the Hamiltonian formalism had yet to be discovered.
Third, by extending the quantum nature of radiation to mechanical
objects on the ground of their 'kinship', Einstein initiated the quantum
statistical mechanics. Remarkably, it was the success of his theory of speci-
fic heat that convinced people of the reality of quanta in the purely mecha-
nical regime. The Bose-Einstein statistics and the concept of Bose-Einstein
condensation has, a very important place today, not only in statistical phy-
sics but also in elementary particle physics.
In reading through Einstein's paper, we are impressed very much by
the simplicity and ease of his arguments. An example is his proof that a
small part of a microcanonical ensemble is a canonical ensemble. It may
be counted as the fourth of Einstein's contributions that he provided plain
physical interpretations to different problems in statistical thermodynamics.
We may conclude with Born 41 that Einstein is one of the fathers of sta-
tistical mechanics.
While Einstein's development of the statistical theories was insepara-
bly connected with his exploration into the quantum wonderland, we know
that he turned to insist, "God does not play with dice" after the quantum
mechanics was formulated with the probability interpretation of Max Born
as its essential ingredient. 42 Einstein wanted to preserve the rigid causality
of the classical physics and to restrict the use of the probability to the cases
with unmanageably large systems or systems for which the details of the
initial conditions were unknown. Einstein's attitude towards the quantum
mechanics will be discussed elsewhere in this book.

41 See the remark made by Born in the International Conference on Statistical Me-
chanics in 1949. Supp!. Nuovo Cimento vol. VI, Series IX, 1949, p. 296.
42 See M. Born, ref. 2, in particular Einstein's letter to Born quoted therein.
 89

"On the History of the Special Relativity Theory"

Mthur I. Miller*

Page for page Einstein's relativity paper is virtually unparalleled in

the history of science in its depth, breadth and intellectual virtuosity. Setting
the paper in its historic context serves to make this point manifestly clear.
There are striking similarities between the physics of 1905 and that of
1979. For then, as at present, many of the foremost physicists were seeking
a unified field-theoretical description of matter, e.g. Max Abraham, H. A. Lo-
rentz, Henri Poincare and Wilhelm Wien.
However, life was "simpler" in 1905, for they had only one elementary
particle, the electron;1 and there existed a theory which was considered to
be capable, after some further tinkering, to serve as the cornerstone of their
research effort - H. A. Lorentz's electromagnetic field theory. I shall say
a few words about the chief stages of development from 1900-1905 of this
electromagnetic world-picture, i. e. the research effort whose goal was to
deduce all of physical theory from Lorentz's electromagnetic theory,2 so
that we may better appreciate how different was Einstein's approach to the
electrodynamics of moving bodies.
By 1900 it was becoming clear that attempts to deduce electromag-
netic theory from the laws of mechanics in accord with a mechanical world-
picture paled before the successes of Lorentz's electromagnetic field theory,
formulated in 1892 - for example, the discovery in 1897 of Lorentz's
electron, Lorentz's explanation of 1896 of the Zeeman effect, and his syste-
matic explanation in 1895 of the results of experiments accurate to first
order in vic concerning the optics of moving bodies, among them are obser-
vations of stellar aberration by Bradley, Arago and Airy, the experiments
of Mascart and Jamin, Hoeck, and Fizeau.

I gratefully acknowledge support for this research from the National Science F ounda-
tion
This essay is based upon my paper "The physics of Einstein's relativity paper of
1905 and the electromagnetic world picture of 1905," American Journal of Physics,
45,1040-1048 (1977)
2 For an in-depth discussion of this research effort see A.I. Miller, "A Study of Henri
Poincare's 'Sur la dynamique de l'electron,' " Archive for History of Exact Sciences,
10, Nos. 3-5, 207 328 (1973)
90 Arthur I. Miller

Equations (1) to (5) are the fundamental equations of Lorentz's electro-

magnetic field theory as published in his seminal pap'er of 1892 "Maxwell's
Electromagnetic Theory and Its Application to Moving Bodies:"3,4

~ ~
'iJ·E=47rp (1)
~ ~
'iJ·B =0 (2)
~ Maxwell-Lorentz
~ ~ 1 aB Equations
'iJXE=---
c at (3)
~
~ ~ 1 aE 47r ~
'iJXB=--+-pv
c at C
(4)

Lorentz force
(5)
equation

where p is the electron's charge density, c = 3 . 108 m/sec.

These equations are mathematically equivalent to the ones used today.
Lorentz considered the fundamental equations as axiomatic, i. e. he made
no attempt to derive them from the laws of mechanics. The sources of the
electromagnetic field are electrons moving about in an all-pervasive stagnant
ether. The fundamental equations have the form of Eqs. (1) to (5) only
relative to an ether-fixed reference system, for according to ether-based theo-
ries of l~ht, only relative to such a system is the velocity of light exactly
C; thus, v is the electron's velocity relative to the ether. In short, Lorentz's
ether-fixed reference systems are preferred reference systems.
Lorentz, in his classic monograph of 1895 entitled "Treatise on a
Theory of Electrical and Optical Phenomena in Moving Bodies~',s explained
the negative results of the first-order optical ether-drift experiments in neu-
tral, nonconducting nonmagnetic matter in a way that can be described as
follows:

H. A. Lorentz, "La theorie electromagnetique de Maxwell et son application aux

corps mouvants," Arch. neerl., 25, 363 (892); reprinted in Collected Papers (9
Vols., Nijhoff, The Hague, 1935-1939), Vol. 2,164-243.
4 For discussions of Lorentz's seminal paper of 1892 and for further references see
Ref. 2 and A. I. Miller, "On Lorentz's Methodology," The British Journal for the
Philosophy of Science, 25, 29-45 (1974).
5 H. A. Lorentz, Versuch einer Theorie der elektrischen und optischen Erscheinungen
in bewegten Korpern Ost ed., Brill, Leiden, 1895; 2nd ed., Teubner, Leipzig, 1906).
See Refs. 2 and 4 for further discussions of this monograph. All references will be to
the second edition which is unchanged from the first.
On the History of the Special Relativity Theory 91

The Maxwell-Lorentz equations relative to an ether-fixed reference

system S and in a charge-free region are
--+--+
'iJ . E =0 (6)
--+--+
'iJ . B =0 (7)

--+
'iJ X E
--+
=-
aB
--+
c1 at (8)

Consider that the Galilean transformation is modified to be

Xr = X - vt (10)

Yr = Y (11)

Zr = z (12)

v (13)
tL = t -'2 x
c '
where x, y, z and t are the spatial and temporal coordinates in S; X r , Yr ,
Zr are the spatial coordinates in an inertial reference system Sr moving
with velocity v relative to S and along their common x-axes; and tL is the
"local time" coordinate 6 which is a nonphysical, i. e. mathematical time
coordinate - the real physical time in Sand Sr is the Galilean time t. Lo-
rentz discovered that if the Maxwell-Lorentz Equations (6) to (9) are trans-
formed to Sr using the set of modified Galilean transformations, and appro-
priate transformation equations for the electromagnetic field quantities
are used, viz.,
--+
--+
Er = E +
--+
cv X --+B (14)
--+
--+
Br =B
--+
-cv X --+E, (15)

then to first order in vic the Maxwell-Lorentz equations in Sr have the

same form that they have in S (see Eqs. (6) to (9», i.e.
--+ --+
'Vr . Er = 0 (16)
--+ --+
'Vr . Br =0 (17)

(18)

(19)

6 Ibid, P 49.
92 Arthur I. Miller

where ~r = (a~r' atr ' a t ) . This approximate convariance was referred

to by Lorentz as "the theorem of corresponding states."7 Thus, to first

order in vic phenomena concerning the optics of moving bodies occur on
the moving earth as they do relative to an ether-fixed reference system, i. e.
to this order of accuracy the velocity of light in Sr is c. However, Lorentz
could not explain systematically the result of the interferometer experiment
of Michelson and Morley, an experiment accurate to second order in vic;
rather, the ad hoc Lorentz contraction hypothesis was necessary.8 This
blemish on Lorentz's theory was subjected to epistemological-physical
criticism by Henri Poincare. 9
In fact, Lorentz's successes in the electrodynamics and optics of moving
bodies were achieved at the expense of using a set of space and time trans-
formation equations under which Newtonian mechanics was not covariant.
In other words, Lorentz's theory violated the Newtonian principle of rela-
tivity. Thus, there was presumed to be an incompatibility between the
laws of mechanics and of electromagnetism - something had to give.
In 1900 Wilhelm Wi en proposed that perhaps research should proceed
toward an electromagnetic world-picture in which the laws of mechanics
would be deduced from the laws of Lorentz's electromagnetic field theory.l0
One implication of such a research effort is that since the electron's mass
should arise from its self-fields, then the mass is a function of the electron's
velocity through the ether. Lorentz in 1899, had already speculated upon
the nonconstancy of the electron's mass. l l
Events then began to move rapidly: Walter Kaufmann of G6ttingen,
in a series of experiments beginning in 1901, studied the behavior in parallel
electric and magnetic fields of high velocity electrons from a radium bromide
source. He found that the electron's mass is indeed a velocity dependent
function increasing without limit as the electron's velocity approaches that

7 Ibid., p. 85.
8 See Ref. 4 for further discussion of the ad hocness of Lorentz's contraction hypothe-
SIS.
9 Poincare's criticisms of Lorentz's electromagnetic theory during the years 1895 -1904
are discussed in Part 5 of Ref. 2.
10 W. Wien, ,.Ober die Moglichkeit einer elektromagnetischen Begriindung der Mecha-
nik," Recueil de travaux offerts par les auteurs a H. A. Lorentz (Nijhoff, The Hague,
1900), pp. 96-107.
11 H. A. Lorentz, "Theorie simplifiee des phenomenes electriques et optiques dans des
corps en mouvement," Versl. Kon. Akad. Wetensch. Amsterdam, 7, 507 (1899); re-
printed in Collected Papers, Vol. 5, 139-155.
On the History of the Special Relativity Theory 93

of light. In 1902-1903 Kaufmann's colleague at Gottingen Max Abraham

formulated the first field-theoretical description of an elementary particleY
Abraham's electron is a rigid sphere whose mass, entirely of electromagnetic
origin, is a two-component function of the electron's velocity. Abraham's
predictions for the electron's longitudinal (md mass and transverse (mT)
mass are

m
L
= m c2
0 v2
l' 2 vic _ 10 (1 + vic )]
v2 g \ 1 - vic (20)
] --
(2

m
T -
- mo C 33 .1 11 + V 22 ) log
2 v '- \ c /
(11 -+ ViC)
vic
- 2 vic Jl (21)

2
where mo = 8~ is the electron's electrostatic (i.e. rest) mass, and r is the
7r rc
electron's radius. The transverse mass agreed with Kaufmann's data. The
goal of the electromagnetic world-picture appeared to have been reached.
To be sure, Abraham's theory of the electron could not explain the
results of the high precision second-order ether-drift experiments of Lord
Rayleigh in 1902 13 and D.B. Brace l4 in 1904 which sought unsuccessfully
to detect double refraction in moving isotropic bodies. Nevertheless, Abra-
ham believed that in time his theory of the electron could be slightly altered
to explain optical experiments as well.
In 1904, Lorentz, prompted by the criticisms of Poincare and the
appearance of new data accurate to second order in vic (i.e. in addition
to the data of Michelson and Morley, the data of Rayleigh and Brace and
of Trouton and Noble) IS, extended the application of his electromagnetic
theory to a theory of the electron; Lorentz's now famous paper of 1904 is
entitled "Electromagnetic phenomena in a system moving with any velo-

12 M. Abraham, "Dynamik des Elektrons." Nachr. Ges. Wiss. Gottingen, 20-41 (1902);
"Prinzipien der Dynamik des Elektrons," Annalen der Physik, 10, 105-179 (1903).
See Part 3 of Ref. 2 for further discussion of Abraham's theory of the electron.
13 Lord Rayleigh, "Does Motion through the Aether cause Double Refraction?," Phil.
Mag .. 4, 678-683 (1902)
14 D. B. Brace, "On Double Refraction in Matter moving through the Aether," Phil.
Mag., 7.317- 329 (1904)
15 Trouton and Noble attempted unsuccessfully to measure the turning couple of a
condenser which was suspended from a thread at a fixed point on the earth. See
F. T. Trouton and H. R. Noble, 'The Mechanical Forces Acting on a Charged Electric
Condenser Moving through Space," Phil. Trans. Roy. Soc. London A, 202, 165-181
(1903)
94 Arthur I. Miller

city smaller than that of light. "16 Lorentz's electron is deformable, under-
going a Lorentz contraction when in motion (an hypothesis that Abraham
assiduously avoided). Lorentz claimed that his theory could explain the
negative results of ether-drift experiments accurate to all orders in vic.
Lorentz's predictions for the electron's longitudinal (mL> and transverse
(mT) masses are:

4 (22)
mL =-
. 3

(23)

2
where mo = __e_2 is the electron's electrostatic (i. e. rest) mass, and r is
81Trc
the electron's radius. Furthermore, Lorentz claimed that his prediction
for the electron's transverse mass also agreed with Kaufmann's data, within
the experimental error.
But Henri Poincare, at the Congress of Arts and Science in St. Louis
Missouri in 1904, praised Lorentz's new theory of the electron.'7 He empha-
sized that now the Lorentz contraction hypothesis, originally cooked up for
the purpose of explaining one experiment, was no longer ad hoc since it
could now explain several second-order experiments and was folded into the
new theory of the electron as one of several "complementary hypotheses. "18
Poincare referred to Lorentz's new generalized theorem of corresponding
states embodied in his 1904 paper as the "principle of relativity" which
asserts that "the laws of physical phenomena must be the same for a station-
ary observer as for an observer carried along in a uniform motion of trans-
lation."19 Poincare also realized that further work was required on Lorentz's
theory of the electron.

16 H. A. Lorentz, "Electromagnetic Phenomena in a system mov:ng with any velocity

smaller than that of light," Proc. Roy. Acad. Amsterdam, 6, 809 (1904); reprinted
in The Principle of Relativity: A Collection of Original Memoirs on the Special and
General Theory of Relativity by H. A. Lorentz, A. Einstein, H. Minkowski and H.
Weyl, translated by W. Perrett and G. B. Jeffery and notes by A. Sommerfeld (Dover
Publications, New York, n.d.) pp. 11-34. Hereafter this reprint volume will be re-
ferred to as PRe. See Refs. 2 and 4 for further discussion of this paper.
17 H. Poincare, "L'etat actuel et l'avenir de la Physique mathematique," delivered on
24 September 1904 at the International Congress of Arts and Science at St. Louis,
Missouri and reprinted in H. Poincare The Value of Science, translated by G. B.
Halsted (Dover Publications, New York, 1958), pp. 91-111.
18 Ibid., p. 100.
19 Ibid., p. 94.
On the History of the Special Relativity Theory 95

There appeared in the Camptes Rendus of 5 June 1905 a first short

version of a major paper by Poincare which would appear in full elsewhere
the following year; both versions bear the title "On the Dynamics of the
Electron."2o Here Poincare corrects certain technical errors in Lorentz's
paper of 1904 and goes on to prove that of all possible models of the elec-
tron wherein the electron's mass is generated by its self-fields, only Lorentz's
is consistent with the principle of relativity. To obtain this result Poincare
found it necessary to add an additional term to the Lagrangian for the
electron's self-fields. He interpreted this term as the energy due to a stress
of unknown origin internal to the electron. One function of this stress
(which became known as the Poincare stress) was to bind Lorentz's deform-
able electron; thus, responding to criticisms of Abraham made in 1903
through 1905 that Lorentz's deformable electron would explode. 21 Poin-
care's is a beautiful and classic paper and contained for the first time many
mathematical techniques which would eventually find their physical inter-
pretation in the special relativity theory, as a result of the researches of
Hermann Minkowski of Gottingen - for example, the seeds of the four-
vector formalism, the notation x, y, z, ict to extend the result of the theory
of invariants in a three-dimensional Euclidean space to four dimensions,
the use of Lorentz invariance and covariance toward the formulation of
physical theories, in particular a theory of the electron and a theory of
gravitation. 22
To summarize thus far: By 1905 many physicists considered Lorentz's
theory of the electron to be the cornerstone of a unified field-theoretical
description of nature. It was a theory in which the electron's kinematics
were deduced from its dynamics, and in which effects had dynamical causes
- for example, because of the interaction with the ether of the bound
electrons that constitute a macroscopic body, the body will undergo a
Lorentz contraction, its mass will be a function of velocity, and thus is
explained why the velocity of light when measured in an inertial reference
system always turns out to be c in all directions. On the negative side of the

20 H. Poincare, "Sur la dynamique de I'electron," Comptes rendus de l'Academie des

Sciences, 140, 1504-1508 (1905); Rend. del. Circ. Mat. di Palermo, 21,129-175
(1906) and submitted 23 July 1905. See Ref. 2 for an in-depth study of this classic
paper.
21 See Ref. 12, M. Abraham, "Die Grundhypothesen der Elektronentheorie," Phys. Z.,
5, 576-579 (1904) and M. Abraham, Theorie der Elektrizitiit (2 vols.; Teubner,
Leipzig, 1905), Vol. 2, esp. pp. 205 ff. See Part 4.7 of Ref. 2 for discussion of Abra-
ham's criticisms.
22 Minkowski expressed his debt to Poincare in H. Minkowski, "Das Relativitiitsprinzip,"
delivered 5 November 1907 before the Math. Ges. Gottingen and published in Anna-
len der Physik, 47, 927-938 (1916). See Part 7.1 of Ref. 2 forfurther discussion of
Poincare's influence on Minkowski.
96 Arthur I. Miller

ledger, Lorentz's theory was heavy with hypotheses, including the new
spacetime transformations and the Lorentz contraction; but they seemed
necessary in order to explain the experimental data - especially the meas-
ured value for. the velocity of light. The principle of relativity of Lorentz
and Poincare has the aura of having been induced from the experimental
data, in particular that of Michelson and Morley. Also, it was perceived
that some (presumably minor) elaborations remained to complete the
theory - for example, Newton's second law could only be deduced ap-
proximately from the Lorentz force law because terms in the expansion
of the electron's self-fields containing derivatives of the electron's acceler-
ation had to be dropped. 23
Mechanics appeared to be becoming a branch of electromagnetism that
was consistent with a principle of relativity. The stage was set. A great new
era in science was emerging and the main actors to play their parts were
known - or were they?
It is indeed an irony that less than a month after the brief version of
Poincare's "On the Dynamics of the Electron" appeared in the Comptes
Rendus, there arrived at the editorial office of the Annalen der Physik a
manuscript to be considered for publication in its Volume 17 by the author
of two previous papers in that volume - Albert Einstein. The paper is en-
titled "On the Electrodynamics of Moving Bodies. "24
To further set the stage for our discussion of Einstein's paper I shall
review briefly what we know of his pre-1905 knowledge of electrodynamics.
From a variety of mutually confirming sources we know that Einstein
while a student at the ETH in Zurich from 1896 to 1900 had studied Lo-
rentz's seminal paper of 1892 on electromagnetic theory and his treatise on
electrodynamics of 1895; and furthermore, that Einstein had seen neither
Lorentz's paper of 1904 nor Poincare's "On the Dynamics of the Elec-
tron. "25 One can also substantiate that Einstein before 1905 had studied
Abraham's theory of the electron and Abraham's criticisms of 1904 on

23 For details see Part 6.8 of Ref. 2.

24 A. Einstein, "Zur Elektrodynamik bewegter Korper," Annalen der Physik, 17,
891-921 (received 30 June 1905); reprinted in PRe, pp. 37-65.
25 See G. Holton, "On the Origins of the Special Relativity Theory," in G. Holton,
Thematic Origins of Scientific Thought: Kepler to Einstein (Harvard University
Press, Cambridge, 1973), pp. 165-183. Hereafter this book will be referred to as
Thematic Origins and all references to Holton's papers will be to Thematic Origins.
G. Holton, "Influences on Einstein's Early Work," in Thematic Origins, pp. 197-217.
See M. Born, Physics in My Generation (Springer, New York, 1969), esp. p. 104 for a
translation of a letter from Einstein to his biographer C. Seelig; this translation ap-
pears also in Ref 2 p. 268.
On the History of the Special Relativity Theory 97

Lorentz's new theory of the electron. 26 Einstein said later in his "Autobio-
graphical Notes" of 1946 that he had perceived the instability of Lorentz's
electron as a "fundamental crisis, "27 yet he had in 1900 already discerned
an even greater crisis; namely, that accepting Planck's radiation law as valid
vitiated the possibility of either an electromagnetic or mechanical world-
picture. 28
Let us now turn to Einstein's relativity paper and compare his approach
to the electrodynamics and optics of moving bodies with that of the ether
theorists.
The relativity paper is structured thus: an untitled introductory section,
a Part I entitled "Kinematical Part," and a Part II entitled "Electrodynam-
ical Part." Each part is subdivided into five sections. The order of the subject
matter is thus the opposite of papers written by proponents of the electro-
magnetic world-picture, who emphasized the dynamics of the electron, i.e.
they deduced the electron's kinematics from its dynamics.
It is on the first page of the introductory section, a section that is
perhaps the most familiar, that Einstein sets the theme in the opening sen-
tence: 29 "It is known that Maxwell's electrodynamics - as usually under-
stood at the present time - when applied to moving bodies, leads to asym-
metries which do not appear to be inherent in the phenomena." It can be
established that Einstein is referring here to the Maxwell-Lorentz theory,
i.e. Lorentz's electromagnetic theory, and not to Maxwell's original formu-
lation. 30 Einstein is not asserting that there are errors in the formulation
of the Maxwell-Lorentz theory, rather that it is misinterpreted; further-
more, the misinterpretation "leads to asymmetries which do not appear to
be inherent in the phenomena." As an example of an asymmetry Einstein in
his well-known first paragraph discusses, as a gedanken experiment, the cur-
rent that arises in a conducting loop that is in inertial motion relative to a
constant magnetic field. Although the magnitude and direction of the in-
duced current depends only upon the relative velocity of conductor and
magnet, the Maxwell-Lorentz theory, Einstein explains, "as understood at

26 See A. L Miller, "On Einstein, Light Quanta, Radiation and Relativity in 1905,"
American Joumal of Physics, 44, 912-923 (1976).
27 A. Einstein, "Autobiographical Notes," in P. A. Schilpp (ed.), Albert Ei1stein: Philos-
opher-5cientist (The Library of Living Philosophers, Evanston, 1949), p. 37.
28 Ibid., pp. 52-53.
29 Ref. 24, p. 37.
30 E.g Einstein in his essay "Relativitatsprinzip und die aus demselben gezogenen
Folgerungen," Jahrbuch der Radioaktivitat und Elektronik, 4, 411-462 (1907)
refers to the set of Eqs. (1) to (4) as the "Maxwell-Lorentz equations" (p. 427),
and discusses only "H. A. Lorentz's electrodynamics of moving bodies" (p. 412).
98 Arthur I. Miller

the present time ... draws a sharp distinction' '31 between the cases of
magnet in motion and conductor at rest and vice versa.
Did others feel there was an asymmetry in the contemporaneous anal-
ysis of the case of magnet and conductor? After all Einstein began the rela-
tivity paper with the phrase "It is known ... ". However the works of Max-
well, Hertz, Lorentz and Poincare reveal no discussion of asymmetries in
electromagnetic induction.
Einstein then begins the next paragraph, which is breathtaking in
scope: 32 "Examples of this sort, together with the unsuccessful attempts
to discover any motion of the earth relatively to the "light medium," suggest
that the phenomena of electrodynamics as well as of mechanics possess no
properties corresponding to the idea of absolute rest." Einstein's linking
of the experiment of magnet and conductor to an apparently unrelated
set of experiments - the ether drift experiments (left unnamed) - is one
of the master strokes to be found in the relativity paper. For Einstein
realized the importance of the fact that the theoretical interpretation of the
case of magnet and conductor discussed in the first paragraph, depends upon
the laws of mechanics and electromagnetism. However, according to New-
ton's mechanics there are no preferred inertial reference systems; further-
more, the Newtonian principle of relativity obtains for such systems.
Einstein continues: 33 "They suggest rather that, as has already been
shown to the first order of small quantities, the same laws of electrodynam-
ics and optics will be valid for all frames of reference for which the equa-
tions of mechanics hold good."
Thus, since the ether drift experiments, both electrodynamical and
optical, were unsuccessful - especially important to Einstein are the ones
accurate to first order in vic - and since the Lorentz theory unifies both
electricity and optics, then to first order in vic the laws of mechanics and
electromagnetism should be the same in every inertial reference system.
Then, Einstein boldly widens the Newtonian principle, of relativity to em-
brace both mechanics and electromagnetism: 34 "We will raise this conjec-
ture (the purport of which will hereafter be called the "Principle of Rela-
tivity") to the status of a postulate .... "

31 Ref. 24, p. 37.

32 Loc. cit.
33 Ibid., pp. 37-38. The asterisk refers to a note by Sommerfeld. Unfortunately the
editors did not distinguish between Einstein's and Sommerfeld's notes. There were
only four notes by Einstein in the 1905 relativity paper and no references. The notes
in the Annalen der Physik (designated as AP) and their translations in the Dover
reprint volume (designated as D) are: (AP = 839, D = 39);(AP = 896, D = 42); (AP =
903, D = 48);(AP = 909, D = 53).
34 Ibid., p. 38.
On the History of the Special Relativity Theory 99

Einstein next posits a second axiom: "namely, that light is always

propagated in empty space with a definite velocity c which is independent
of the state of motion of the emitting body. "35 In the Maxwell-Lorentz
theory this statement is axiomatic only in ether-fixed reference systems and
not in inertial reference systems. According to the Maxwell-Lorentz theory
these two sets of reference systems are related by a transformation that is
mathematically equivalent to the Galilean transformation, i.e. since the
ether-fixed reference systems never move, the group structure of the Gali-
lean transformations has no physical meaning. 36 However, according to the
Einstein principle of relativity there are no preferred inertial reference sys-
tems, thereby rendering Lorentz's ether as "superfluous."37 Therefore, the
light axiom holds in all inertial reference systems. Einstein has thus eliminat-
ed one more asymmetry (recall that he used the plural form of the word in
the opening sentence).
Einstein, in Section 2 of the relativity paper, will restate the principle
of relativity as follows: 38 "The laws by which the states of physical systems
undergo change are not affected, whether these changes of state be referred
to the one or the other of two systems of coordinates in uniform translatory
motion." Poincare's and Einstein's principles of relativity, although worded
similarly, are fundamentally different. (Incidentally, the term "principle
of relativity" was widely used c. 1900, especially in discussions on the foun-
dations of geometry.39) Poincare's principle of relativity was a generalized
statement of Lorentz's theorem of corresponding states, and hence part of
a well-received reductionistic research effort. Poincare's principle of relativi-
ty was not an axiom (or a convention 40 ) and was closely linked with rela-
tively recent experimental data, especially that of Michelson and Morley.
With Einstein the case was different. Historical scholarship - in partic-
ular studies by Gerald Holton - have shown that although experimental
data of course played a role in Einstein's thought toward a principle of
relativity, they, and particularly the puzzles of (V!c)2 measurement, did not

35 Loc. cit.
36 According to Lorentz's theory the physical coordinate transformation is the Galilean
transformation - for example the local time coordinate in Eq. (13) is a further mathe-
matical scale transformation on the Galilean time, i. e. the absolute time. But since
the ether-fixed reference system S never moves, the inverse transformation has no
physical meaning. See note 56 below for further discussion.
37 Ref. 24, p. 38.
38 Ibid., p. 41.
39 See Ref 2, note 62 on p. 233.
40 For further discussion of Poincare's philosophy of science and its influence on the
direction of his scientific researches, see Ref. 2 and A. I. Miller, "Poincare and Ein-
stein: A Comparative Study," to be published in Volume XXXI of the Boston Stu-
dies in t he Philosophy of Science.
100 Arthur I. Miller

decide the issue for him.41 Only a few, old, quite well established and seem-
ingly well understood experiments would be needed. To Einstein the
first-order experiments concerning stellar aberration and Fizeau's measure-
ments of the velocity of light in moving water "were enough" (as Einstein
once said himself).42 They had in fact been explained systematically by
Lorentz in his treatise of 1895, using the hypothesis of an unphysical local
time coordinate. Einstein later wrote that before 1905 it had occurred to
him that perhaps one should define as time what Lorentz in the treatise
of 1895 had considered merely an auxiliary mathematical hypothesis. 43
But the mathematics would come later, for Einstein did not think in such
terms; rather, there was the predominantly visual component to his thought
from which arose some of his ideas on the nature of light that would lead
to the special relativity theory.44 In particular there was the gedanken ex-
periment of magnet and conductor, and the gedanken experiment from
the Aarau period of 1895 wherein Einstein considered what it would be like
to run alongside of a light wave and then catch up with a point on it. Ein-
stein wrote of the Aarau experiment in his "Autobiographical Notes" of
1946 that it "was intuitively clear"45 that everything should occur for the
moving observer as it did for an observer at rest relative to the earth. He
continues: "For how, otherwise, should the moving observer be able to
determine that he is in a state of fast uniform motion?"46 Einstein consid-
ered this as a paradox containing, as he wrote in 1946, with hindsight,
"the germ of the special relativity theory. "47 But first says Einstein in the
"Autobiographical Notes" he had to realize that the "absolute character
of time, viz., of simultaneity unrecognizedly was anchored in the uncon-
scious." "The type of critical reasoning," Einstein continues, necessary for
this discovery he found in the philosophical works of Hume and Mach. 48 In
fact, by 1905 Einstein may very well have realized that this gedanken
experiment contained in principle every optical ether-drift experiment
that could be performed. Thus, Einstein was working on the same problem
that was of concern to Lorentz and Poincare; namely, why does the veloc-

41 See G. Holton, "Einstein, Michelson and the 'Crucial' Experiment," in Holton,

Thematic Origins, pp. 261-352.
42 R. S. Shankland, "Conversations with Albert Einstein," American Journal of Phys-
ics, 31, 47-57 (1963), p. 48.
43 Ref. 30, p. 413.
44 For a discussion of the visual component in Einstein's thought see G. Holton, "On
Trying to Understand Scientific Genius," in Holton, Thematic Origins, pp. 353-380.
45 Ref.27,p.53.
46 Loc. cit.
47 Loc. cit.
48 Loc. cit.
On the History of the Special Relativity Theory 101

ity of light measured in an inertial system seem always to be c? Lorentz

tried to explain this effect, known to be accurate to second order in vic,
with dynamical causes and a mathematical time coordinate, the local time.
Einstein, on the other hand, keeping in mind the first order experiments,
the gedanken experiments, the philosophical writings of Hume and Mach,
and as Martin J. Klein has emphasized, the laws of thermodynamics,49
boldly moved counter to the prevailing currents of theoretical physics
to resolve this problem in a Gordian manner; namely, by setting up a phys-
ics in which such problems do not appear. Einstein's two axioms of rela-
tivity theory do not explain the failure of the ether-drift experiments, or
equivalently why the measured velocity of light always turns out to be c,
and why one cannot catch up with a light wave. Rather, it is by definition
that these experiments must fail, and it is by definition that the space of
every inertial reference system is homogeneous and isotropic for the propa-
gation of light. Einstein makes use of this property of space many times
in the relativity paper.
In t he concluding paragraph to the introductory section Einstein
further qualifies the theme already introduced in the example of mag-
net and conductor. Einstein explains that in his opinion the difficulties
which the electrodynamics of moving bodies encounters is rooted in kine-
matics and not dynamics, i. e. "relationships between rigid bodies (systems
of coordinates), clocks, and electromagnetic processes."50 Furthermore,
by introducing at this point the rigid body as an irreducible entity, Einstein
makes clear his intent not to speculate on the constitution of matter. Ein-
stein recalled in his Nobel Prize address of 1923 that the time was not yet
ripe in 1905 for such speculations, and that on this point the physicists of
1905 were out of their "depth."51
I next turn to the Section 1 of the relativity paper entitled "Definition
of Simultaneity." Einstein begins by presenting definitions that are in
principle operational of the concepts inertial reference system and position
relative to an inertial reference system. The definitions are based upon "the
employment of rigid standards of measurement and the methods of Euclid-
ean geometry." 52 Almost certainly the reason why Einstein considered
it important that these definitions be operational, at least in principle, is
the vague and inconsistent manner in which these concepts were discussed

49 See, for example, M. J. Klein, "Thermodynamics in Einstein's Thought," Science,

157,509-516 (J 967)
50 Ref. 24, p. 38.
51 A. Einstein, "fundamental Ideas and Problems of the Theory of Relativity," deliv-
ered 1 I July 1923 to the Nordic Assembly of Naturalists at Gothenburg in ac-
knowledgement of the Nobel Prize; in Nobel Lectures: 1901-1921 (Elsevier, New
York, 1967), pp. 482-490, esp. p. 484.
52 Ref 24. p. 38
102 Arthur I. Miller

in ether theories. There, knowledge of the true length of a body was appar-
ently forever denied because one could not determine the earth's velocity
relative to the ether. Einstein then explains that the coordinates of a material
point in motion are functions of time and thus we must understand what
we mean by "time."53
Einstein goes on to argue that simultaneity is not an absolute concept;
rather, there should be a distinction between local and distant simultaneity.
Einstein is careful not to let sensations decide the issue as did the mecha-
nists, the ether theorists, and philosopher-scientists such as Mach and Poin-
care ;54 rather, an in principle operational definition is necessary to define
distant simultaneity and time. Einstein obtains a definition of clock syn-
chronization, and hence time for clocks at relative rest in any inertial refe-
rence system, by means of an in-principle-operational procedure in which is
embodied the homogeneity and isotropy of space for the propagation of
light. It is a two-way two-clock light exchange experiment. Consider two
clocks at relative rest and at positions A and B in an inertial reference sys-
tem. Light is emitted from the clock at A at time tA , received at the clock
at B at time tB where it is reflected back to A arriving at time tA'. Einstein
is led to the definition of the time tB as one-half the sum of the times tA
and tA', i.e.
tA + tA'
tB = 2 (24)

Einstein begins Section 2 entitled "On the Relativity of Lengths and

Times," by restating the two principles or axioms of relativity and then
demonstrates roughly, using the definition of simultaneity from Section 1,
i. e. Eq. (24), that length and time are relative quantities. Thus, in contrast
to the theory of Lorentz, there are no true lengths, and time is not an abso-
lute quantity.
The remainder of the relativity paper has the pristine form of a geo-
metry treatise: all results flow from the two axioms of relativity theory
and the definition of simultaneity proposed in Section 1. It is as if Ein-
stein had written the relativity paper in the style of another major treatise
on mechanics, published 218 years earlier, and which was consciously styled
after Euclid's Elements - namely, Newton's Principia. 55
Einstein, in Section 3 entitled "Theory of the Transformation of Co-
ordinates and Times from a Stationary System to another System in Uni-
form Motion of Translation Relatively to the Former", deduces the rela-
tivistic transformation equations for the space and time coordinates. To

53 Ibid., p. 39.
54 See, for example, my essay "Poincare and Einstein ... " in Ref. 40.
55 For further comparisons of the styles in Einstein's relativity paper and Newton's
Principia, see Holton's "On the Origins ... " in Ref. 25, esp. pp. 170-171.
On the History of the Special Relativity Theory 103

Poincare and Lorentz these transformations were separate hypotheses, in

which all symbols did not have a precise operational meaning due to un-
known velocities relative to the ether. However, in Einstein's theory there
are no unknown or unknowable quantities - spatial positions relative to
an inertial system are measured with rigid rods and the time at each point
is read from a clock at that point at rest relative to the inertial system and
synchronized according to the definition from Section 1 (Eq. (24». I shall
skip a couple of pages to Einstein's well-known relativistic transformation
equations:

I
x - vt
(25)
X = ( v2 ) 1/2
1--
c2

y' =y (26)
,
z =z (27)
v
t -- X
, c2
t =, 2)112 ' (28)
( 1- ~
(2 ,

where the primed (unprimed) space and time coordinates are measured rela-
tive to the inertial reference system k (K).s6 Einstein interprets them as relat-
ing between two inertial reference systems the readings on clocks and meas-
uring rods. (Unknown to Einstein a mathematically equivalent set of these
equations had appeared in Lorentz's paper of 1904 and had been dubbed
by Poincare in 1905 the Lorentz transformations. s7 )
In Section 4 entitled "Physical Meaning of the Equations Obtained
in Respect to Moving Rigid Bodies and Moving Clocks" Einstein deduces
the apparent contraction of bodies in inertial motion as measured by an
observer in another inertial reference system, and the equation for time
dilation. To Lorentz and Poincare the length contraction was a separate
hypothesis. Furthermore, in their theory the contraction hypothesis had
no clear physical meaning due to the lack of operational methods to deter-
mine changes of length - for example, congruence methods fail because
all bodies contract and experiments using light signals were considered as
equivalent to a Michelson-Morley experiment. In addition, there exist
in the ether theory true lengths and velocities relative to the ether that
are not experimentally detectable. The phenomenon of time dilation had
no operational meaning at all in the ether theories where time was an ab-
solute quantity.58

S6 See Refs. 2 and 4 for a comparison of Einstein's transformations with Lorentz's.

S7 Ref. 20, p. 130 of the paper of 1906.
S8 For further discussion see Refs. 2 and 4.
104 Arthur I. Miller

In Section 5 entitled "The Composition of Velocities" Einstein de-

duces the new addition law for velocities and then proves that the rela-
tivistic transformations form a group. In the theory of Lorentz and Poin-
care the former result contained unknown velocities and the group pro-
perty of the relativistic transformations had no physical meaning. S9
This completes Part I.
Next, in Part II, Einstein applies the new kinematics to electrodynam-
ics. The first Section is numbered Section 6 and is entitled "Transforma-
tion of the Maxwell-Hertz Equations for Empty Space. On the Nature of
the Electromotive Forces Occurring in a Magnetic Field During Motion."
There Einstein begins discussing the properties of radiation in free space
by analyzing what he refers to as the Maxwell-Hertz equations without
source terms. These are actually the Maxwell-Lorentz equations (Eqs. (1)
to (4) with p = 0) and in his review paper of 1907 Einstein uses that name
for them. 30, 60 Einstein considers them to be axiomatic. Then, by demand-
ing that they be covariant - consistent with the two axioms of the special
relativity theory - Einstein deduces new laws of physics; namely, the rela-
tivity of the electromagnetic field quantities. In the electron theory of Lo-
rentz and Poincare covariance is a purely mathematical property interpreted
as transforming the field equations to a mathematical coordinate system
which has all the properties of an ether-fixed reference system. In another
feat of reducing "unnecessary" separate entities, Einstein then demonstrates
that the Lorentz force is not a separate axiom as it had been in Lorentz's
theory. Einstein will demonstrate this point in further detail in his Section
10.
The concluding paragraph to Section 6 is characteristic of the literary
style of this paper. Einstein writes: 61 "Furthermore it is clear that the
asymmetry mentioned in the introduction as arising when we consider
the currents produced by the relative motion of a magnet and a conductor,
now disappears. Moreover, questions as to the "seat" of electrodynamic
electromotive forces (unipolar machines) now have no point." Thus, first
Einstein asserts that together all these results provide the resolution of the
asymmetry in the case of magnet and conductor. He leaves the details as an
exercise to the reader. Then, true to form, Einstein concludes this Section
by audaciously dismissing in one sentence as meaningless a problem consid-

59 See Part 6.5 of Ref. 2 for a discussion of Poincare's analysis of the Lorentz group.
60 Einstein, in 1905, may very well have used the designation Maxwell-Hertz equations
to emphasize his desire not to speculate on the constitution of matter. Hertz's
electrodynamics did not have an atomistic basis; rather, it was an electrodynamics
for matter in bulk. Furthermore Einstein used Hertz's nomenclature for the electric
and magnetic quantities, referring to them as "electric and magnetic forces" (Ref. 24,
p. 52) and giving to them in-principle operational definitions (Ref. 24, p. 54).
61 Ref.24,p.55.
On the History of the Special Relativity Theory 105

ered most perplexing and elaborated upon by all the great electrodynamic-
ists; namely, the determination of the seat of the electromotive force in a
unipolar induction machine.
In Sections 7 and 8 Einstein uses the results of Section 6 and the
new kinematics to solve certain problems concerning radiation in free
space. Invoking the invariance of the phase of a plane wave Einstein, in
Section 7 entitled "Theory of Doppler's Principle and of Aberration,"
solves exactly two problems for which no exact solution could be obtained
in ether-based theories - the optical Doppler effect and stellar aberration.
In Lorentz's theory not only did unknown velocities prevent a complete
solution to these problems, but in addition, even first order observations of
stellar aberration, a phenomenon dependent only upon the relative velocity
of earth and star, required two different dynamical explanations depending
upon whether the effect was observed in a geocentric system or in an ether-
based system. Thus, Einstein has removed yet another asymmetry "not in-
herent in the phenomena."
Einstein begins Section 8 entitled "Transformation of the Energy
of Light Rays. Theory of the Pressure of Radiation Exerted on Perfect
Reflectors," by proving that the ratio of the energy to frequency of a
"light complex", i.e. a light pulse, is invariant. 62 Einstein asserts that this
result is "remarkable."63 Remarkable indeed! For Einstein with a flair for
the understatement, customarily found only in literary works of high distinc-
tion, chose not to refer the reader to the principal result in his first publica-
tion in Volume 17 of the Annalen der Physik. 64 There, Einstein had demon-
strated that light obeying Wien's law of radiation can be described as an en-
semble of independent particles or pulses whose ratio of energy to frequency
is a universal constant ~ namely Planck's constant. Einstein continues in
Section 8 by solving exactly the two old problems of the reflection of light
from a moving perfectly reflecting mirror, and the pressure of light on the
moving mirror. His matter-of-fact solution, in two short paragraphs, belies

62 I discuss in detail this section of Einstein's relativity paper in Ref. 26.

63 Ref. 24. p. 58.
64 A. Einstein. "Dber einen die Erzeugung und Verwandlung des Lichtes betreffenden
heuristischen Gesichtspunkt." Annalen der Physik, 17, 132~148 (1905); translated
by A. Arons and M. B. Peppard in American Journal of Physics 33, 367~374 (1965).
This paper is discussed in detail in M. J. Klein, "Einstein's First Paper on Quanta,"
The Natural Philosopher. 2, 59~86 (1963). Einstein published a third paper in the
volume 17 of the Annalen, "Die von der molekularkinetischen Theorie der Warme
geforderte Bewegung von in ruhenden Fliissigkeiten suspendierten Teilchen," Anna-
len der Physik, 17, 549 (1905); reprinted in A. Einstein, Investigations on the Theory
of the Brownian Movement, translated by A. D. Cowper, with notes by R. Fiirth
(Dover, New York. 1956). Einstein addressed all three papers to a common theme ~
the nature of radiation; I discuss this point in my essay in note 26.
106 Arthur I. Miller

the importance of these problems to the physics of 1905, for they were
considered central to the thermodynamics of radiation. In fact, these pro-
blems could be solved exactly in Lorentz's ether-based theory of electro-
magnetism as Abraham had demonstrated in 1904; however, Abraham's
long path winds torturously through the forty odd pages of his then well-
known paper.6S Furthermore, although the Einstein and Abraham solutions
are mathematically equivalent, they are not observationally equivalent due
to the presence of unknown velocities relative to the ether in Abraham's
work.
In Section 9 entitled "Transformation of the Maxwell-Hertz Equa-
tions when Convection-Currents are Taken into Account" and 10 Einstein
analyzes the so-called Maxwell-Hertz equations for the case including sources.
The sources are point electrons with a totally mechanical mass. The
Maxwell-Hertz equations that Einstein discusses here are mathematically
equivalent to Eqs. (1) to (4); due to the relativity of simultaneity, the
symbols have a different interpretation. Since he is analyzing the motion
of an electron, Einstein is obliged here to finally write that "these equa-
tions are the electromagnetic basis of the Lorentzian electrodynamics
and optics of moving bodies. "66
Einstein, in Section 10 entitled "Dynamics of the Slowly Accelerated
Electron," takes Newton's second law in its familiar mathematical form as
axiomatic in the instantaneous rest system of an electron in an external
electromagnetic field. Then, by transforming to a laboratory inertial refer-
ence system, Einstein deduces the electron's longitudinal (md and trans-
verse (mT) masses. Einstein's results are:

(29)

(30)

where m' is the point electron's mechanical mass. Due to an unfortunate

choice for the definition of force, Einstein's expression for the transverse
mass turned out not to be appropriate, as Max Planck pointed out in 1906. 67

65 M. Abraham, "Zur Theorie der Strahlung und des Strahlungsdruckes," Annalen der
Physik, 14, 236-287 (1904). See Ref. 26 for a discussion of Abraham's paper.
66 Ref. 24, p. 60.
67 M. Planck, "Das Prinzip der Relativitat und die Grundgleichungen der Mechanik,"
Verh. d. p. Ges., 4,136-141 (1906). Until 1911 Einstein's results in the relativity
paper were considered by most physicists as a generalization of Lorentz's theory
of the electron; hence, there was referred to a "Lorentz-Einstein" theory.
On the History of the Special Relativity Theory 107

The result should have been:

I
m (31)
(1- -v2 ')1/2
c2

which is mathematically equivalent to Lorentz's mT in Eq. (23).

Then, Einstein performs the only integral in the entire paper. The
problem is to calculate the kinetic energy of an electron in an external
electrostatic field. The full importance of the solution was realized by Ein-
stein only later in 1905- it turned out to be literally earthshaking. 68
So we have come to the last pages of Einstein's 1905 paper. At the
end, almost as an afterthought, Einstein offers three experiments to test
the new theory. To the best of my knowledge no experimentalist was
prepared to carry out such tests. Indeed, Einstein in yet another bold move
may very well not have listed his prediction for the electron's transverse
mass with the other experimental "tests" for the very reason that it did
not agree with existing experimental data of which he was aware - namely,
Kaufmann's.
The picture that emerges from placing Einstein's 1905 relativity paper
in its historic context is not only that it ran counter to the physics of 1905,
but also that it did not allow the existing data to decide the issue in formu-
lating or in testing the theory.69
The year 1905 was not the end for Einstein, of course. By 1907 he was
already working toward generalizing the special relativity theory, thus

68 Einstein discussed the mass-energy equivalence in the fourth paper he published

in 1905, "Ist die Tragheit eines KCirpers von seinem Energieinhalt abhangig?,"
Annalen der Physik, 18, 630-641 (1905); reprinted in PRe, pp.67-71. Besides
misspelling the title, the Dover edition lists the volume incorrectly as 17.
69 To be sure, I am not asserting that Einstein was disinterested in either experimen-
tal confirmation or in the existence of empirical data. In the case of the genesis
of special relativity theory, as it is throughout the history of science, the interplay
between empirical data or experimental confirmation and the nature of scientific
discovery is not as clear cut as the more positivistically inclined philosophers c.
1905 or c. 1979 would like us to believe. Today this interplay is a central problem
in the history and philosophy of science. See, for example, the essays of Holton
referenced above, my essays in Refs. 4, 40, A. I. Miller, "Albert Einstein and Max
Wertheimer: A Gestalt Psychologist's View of the Genesis of Special Relativity
Theory," History of Science, 13, 75 -103 (1975); and A. l. Miller, "Book Review of
Adolf Griinbaum 's Philosophical Problems of'Space and Time" ISIS 66, 590-594
(1975), Ibid., 68, 449450 (1977).
108 Arthur I. Miller

removing the "logical weakness"7o of the concept inertial reference system

and the aura of effects without dynamical causes. 71 Unlike Newton, Einstein
was to have a second Annus Mirabilis.

70 Ref. 51,p.484.
71 In Part V of Ref. 30 entitled "Relativityprinciple and Gravitation." Einstein gener-
alized the principle of relativity to include accelerating reference systems, and pro-
posed what became known as the principle of equivalence.
 109

Einstein's Model for Constructing a Scientific

Theory
Gerald Holton

1 The epistemological imperative

Judging by his publications and letters, Albert Einstein considered

it one of his important tasks constantly to express and elaborate his views
concerning the philosophy of science. There seem to be two reasons for
that. First, Einstein experienced in his own work in the early years, and
then again among his "ablest students," how important discussions con-
cerning the aims and methods of the sciences are. 1 Such interest was not
merely a matter of intellectual curiosity but, in his opinion, went to the
heart of the task of the innovator: epistemology and science, he said, "are
dependent on each other. Epistemology without contact with science
becomes an empty scheme. Science without epistemology is - insofar as
it is thinkable at all - primitive and muddled." At the same time he warned,
however, that the scientist cannot permit himself to be too restricted "by
the adherence to an epistemological system" (Schilpp, pp. 683-684). He
might therefore seem to be more a philosophical opportunist than school
philosopher. However, that accusation seemed to bother Einstein as little
as did the more serious attacks from so many other quarters upon his science
and his other views.
A second reason why a scientist concerned with the deep problems
should not avoid epistemologic considerations was, in Einstein's opinion,
that there simply was no other way. In our time, when the scientific founda-
tions are changing rapidly, "the physicist cannot simply surrender to the
philosopher the critical contemplation of the theoretical foundations; for
he himself knows best and feels most surely where the shoe pinches" (Ideas
and Opinions, p. 290).
With these motivations, Einstein found himself publishing constantly
on the philosophy of science and, significantly, doing so throughout his
most creative period of scientific work (e.g., 1914: Principles of Theoretical
Physics; 1916: On Ernst Mach; 1918: Motive for Doing Research; 1921:
Geometry and Experience; 1933: On the Method of Theoretical Physics;
1936: Physics and Reality, and many others - not to speak of his letters

1 Cf. A. Einstein, Phys. Zs, 17 (1916), pp. 101 ff.

110 Gerald Holton

to Besso, Solovine, and other friends). With characteristic persistence, not

to say obstinacy, he put on himself the task of presenting what he called
his "epistemological credo." Moreover, it is striking how consistent he was
in his presentation - at least from about 1914 on, after his formative period
during which he had gone through a kind of philosophical pilgrimage.
In the last four decades of his life, he therefore was acting not only as a
profound scientist, but also as a popularizer, teacher, and philosopher-scien-
tist in the tradition of Henri Poincare, Ernst Mach, and others of the genera-
tion before him. It is obvious that he took this role as a public educator
very seriously, and that he tried his utmost to write clearly and at a level
where the intelligent layman would understand him. Hence it came about
that the man who was best known for his legendary struggles with the
most inaccessible and recondite theories in fact was - and to this day re-
mains - one of the most readable and widely read scientists. His essays have
been reprinted in the most distant corners of the world, and of course
there have been volumes upon volumes of analysis of his ideas. It is in this
spirit of Einstein's own intentions that I am glad to respond to the invitation
to provide here an analysis of a key portion of Einstein's epistemological
position, in a manner accessible to a wider audience.

2 Writing to Solovine

In all of Einstein's own writings, one message stands out and is returned
to repeatedly: a model of scientific thinking, and indeed of thinking in gene-
ral. The model forms the core of the first pages of his Autobiographical
Notes,2 which I have analyzed elsewhere. 3 But Einstein's most concise and
graphic rendition of his model is to be found in a letter he wrote to his
friend Maurice Solovine in 1952. I have always thought that for sheer
virtuosity of expression and ability to summarize complex and profound
thoughts, this letter is unique in Einstein's correspondence. It is there-
fore well suited for re-examining his credo, for elaborating on his brief
explanations by reference to others of his publications on the same sub-
ject, and therefore for pulling together many of his methodological ideas
scattered throughout his writings.

2 Einstein wrote the essay in 1946 as the opening article for the book Albert Einstein:
Philosopher-Scientist (Paul A. Schilpp, editor); see Principal References at end of
this essay.
3 G. Holton, "What, Precisely, is 'Thinking'? Einstein's Answer," in A. P. French (ed.),
Centennial volume of the International Commission on Physics Education (in press).
Einstein's Model for Constructing a Scientific Theory 111

Solovine was one of Einstein's oldest friends; they had met in Bern
in 1902, and had kept up a correspondence after Solovine had moved away.
Writing on 25 April 1952, Solovine confesses that he has trouble under-
standing a point made in one of Einstein's essays which Solovine has just
been translating for publication in France. "Would you be so kind," Solo-
vine asks, "as to explain precisely a passage ... which is not quite clear.
You write: The justification (truth content) of the system rests in the
proof of usefulness of the resulting theorems on the basis of sense experien-
ces, where the relations of the latter to the former can only be comprehended
intuitively .... " Solovine indicates his puzzlement and raises questions.
In his reply of 7 May 1952, Einstein starts in his characteristically
relaxed, unpompous manner: "Dear Solo! In your letter you give me a
spanking on the behind ... , but," he continues, "you have thoroughly
misunderstood me with respect to the epistemological matter. Probably
I expressed myself badly." There follows a memorable explanation of the
respective roles of sense experience, intuition, and logic in the functioning
of the imagination. As we shall see, and as one would expect, Einstein
places the emphasis on the sequence of steps in doing science, in making
a discovery or formulating a theory, rather than reformulating the results
later on to make them acceptable to publishers of scientific journals or phil-
osophers interested in the justification of proposed theories.
It will also be noted that while the context of Solovine's question and
Einstein's reply make it clear that Einstein is talking about a model for
thinking in science, nowhere in what follows does he use the word science;
and what fragmentary examples he gives (e.g., relation between the concept
"dog" and the corresponding experiences) are not drawn from scientific
theory. This is entirely in line with his typical refusal to tolerate unnatural
:;lnd unnecessary boundaries. For he said repeatedly that one is dealing here
with a continuum: "Scientific thought is a development of prescientific
thought" (Ideas and Opmions, p. 276); "all this applies as much, and in the
same manner, to the thinking in daily life as to the more consciously and
systematically constructed thinking in the sciences" (Ideas and Opinions
[I.O.], p. 23 ; see also 1.0., p. 324). This point of view was perhaps best caught
in his statement that the "whole of science is nothing more than a refine-
ment of everyday thinking" (1.0., p. 290). Just for that reason, however, the
critical physicist should not restrict his examination of concepts to his own
field of expertise, but should consider "critically a much more difficult
problem, the problem of analyzing the nature of everyday thinking" (ibid.).
Perhaps for this reason Einstein had placed the question "What, precisely, is
'thinking'?" near the beginning of his Autobiographical Notes - and then,
during that discussion, referred only rarely to science.
112 Gerald Holton

3 Given: a "labyrinth of sense impressions"

Einstein begins his explanation to Solovine with the sentence: "I

see the matter schematically thus" - and there follows a diagram (not
surprisingly, for we know of Einstein's preference for visual thinking).
A sketch of great power and simplicity, it concentrates in a few lines a
wealth of information (Fig. 1). The diagram indicates an essentially cyclical

Figure 1

process, and Einstein enters on its discussion by laying out the stage where
the process must both begin and end:
"1. The E (experiences) are given to us."
This refers to the horizontal line shown at the bottom of the figure,
marked E and labeled "Multiplicity [or variety] of immediate (sense) expe-
riences." As usual, these come first in his account, just as he had put the
"reception of sense impressions" as the first item after asking "What is
thinking?" in the Autobiographical Notes. And it will have to come in at the
end also, when we return to the level of sense experience to see if our
theory can handle as large a part of the totality of the facts of experience
as possible - which is, after all, the final test of a theory.
The thin line marked E is rather deceptive. One might better visualize
it as an infinite plane on which the separate and diverse sense experiences
or observations that clamor for our attention are laid out, like so many
separate points. It does indeed represent the "totality of empirical fact"
(1.0., p. 271) or "totality of sense experiences" (Schilpp, pp. 11, 12, and
often elsewhere). In themselves the points on this plane are bewildering,
a universe of elements, a veritable "labyrinth of sense impressions" (1.0.,
p. 291) of which, moreover, we never can be completely sure that they
are not "the result of an illusion or hallucination" (ibid.). In fact, the ulti-
mate aim of science can be defined in this manner: "Science is the attempt
to make the chaotic diversity of our sense-experience correspond to a logi-
Einstein's Model for Constructing a Scientific Theory 113

cally uniform [unified 1system of thought" (1.0., p. 323). The chaotic diver-
sity of "facts" is mastered by erecting a structure of thought on it that
points to relations and order: "In this system, single experiences must be
correlated with the theoretic structure in such a way that the resulting
coordination is unique and convincing" (ibid.).
An aside: Nobody had to point out to Einstein that sense experiences
or "observations" are virtually never pure and unvarnished. Even the father
of positivism, Auguste Comte, had written (Positive Philosophy, 1829) that
without a theory of some sort by which to link phenomena to some prin-
ciples "it would not only be impossible to combine the isolated observa-
tions and draw any useful conclusions, we would not even be able to re-
member them, and, for the most part, the fact would not be noticed by
our eyes." Indeed, sometimes Einstein speaks of "experience" or "facts"
in a way very different from what Ernst Mach took to be "elements";
among facts, Einstein in various writings included the impossibility of
perpetual motion machines, inertial motion, the constancy of light velocity,
and the equality of gravitational and inertial mass (1.0., pp. 307, 309; cf.
letter to Besso, quoted in Holton, Thematic Origins, p. 229). Nevertheless,
in their most primitive form the E in Fig. 1 can be thought of as simple
. .
sensory ImpreSSIOns.

4 The ascent to an axiom system

The diagram in Fig. 1 now goes on to show what is perhaps Einstein's

most insistent conception in epistemology. Rising out of an area just above
a portion of the chaos of observables, E there is an arrow-tipped arch reaching
to the very top of the whole scheme. It symbolizes what under various
circumstances could be a bold leap, a "widely speculative" attempt (ibid.,
p.254), a "groping constructive attempt (ibid., p.286), or a desperate
proposal, made when one has despaired of finding other roads. There, high
above the infinite plane E, is suspended a well-delimited entity labeled "A,
system of axioms ," issuing out of the arrow-tipped arch like a pulse of
light out of the trajectory of a firework.
Einstein writes in explanation:
"2. A are the axioms from which we draw consequences. Psychologi-
cally the A are based upon the E. There is however no logical path from the
E to A, but only an intuitive (psychological) connection, which is always
'subject to revocation.' "
Evidently, Einstein holds that in the formulation of ideas - everyday
as well as scientific ones - the process of thinking or discovery does not
follow the classical model of Mill, i.e., of erecting a logical ladder by induc-
tion of generalizations from the set of individual observations. That method
is only "appropriate to the youth of science" (1.0.,283). Nor does Einstein
114 Gerald Holton

believe, as Ernst Mach had counseled, to remain as much as possible within

the plane of E and to confine oneself to search out the most economic
statements of relations among the elements there; for what that missed,
Einstein explained in his Autobiographical Notes (Schilpp, p.21), was
precisely the "essentially constructive and speculative nature of thought
and more especially of scientific thought."
In the schema of Fig. 1, the arc is just that speculative leap or con-
structive groping to A, the axiom or fundamental principles which in
the absence of a logical path have to be postulated, perhaps at first quite
tentatively on the basis of a conjecture, supposition, "inspiration," "guess,"
or "hunch." We are dealing, after all, with the private process of theory
construction or innovation, the phase not open to inspection by others
and indeed perhaps little understood by the originator himself. But the
leap to the top of the schema symbolizes precisely the precious moment
of great energy, the response to the motivation of "wonder" and of the
"passion of comprehension" (1.0., 342) which can come from the encounter
with the chaotic E. Indeed, there is a clear and uncanny parallel between
the process described in Fig. 1 and the model Einstein proposed to explain
the motivation for research. As Einstein puts it there, to escape from the
chaos of the world of experience, the scientist, scholar, or artist erects a
"simplified and lucid image of the world." lifting into it "the center of
gravity of his emotional life" (quoted in Holton, Thematic Origins, p. 377).
As one would expect from him, Einstein did not speak of the technique
of elevating a supposition or hunch to an axiom or fundamental principle
as if it were some hypothetical advice. He had done so himself in his scienti-
fic papers and, what is more, had confessed it quite frankly. For example,
on the first pages of his first paper on relativity, he refers to a few well-
known experimental facts, some of them in a quite perfunctory manner,
invoking them chiefly to say, without further specification, that they "lead
to the conjecture [Vermutungl" which he calls the Principle of Relativity.
Without further apology or explanation, he then declares "We will raise
this conjecture ... to the status of a postulate." Moreover, he adds at once,
and indeed without any preparation at all, that he will "also introduce an-
other postulate," namely that concerning the constancy of light velocity.
We know that reaching these conjectures, and gathering the courage
to raise them to fundamental principles, were not momentary enthusiastic
decisions but the results of years of groping. It was in fact forced on Einstein
that the kind of fundamental theory he was trying to built could be attained
in no other way. He told in his Autobiographical Notes that it was an act of
desperation when he discovered, shortly after 1900, that the more traditio-
nal way of proceeding, which he later called "Constructive Theories," were
inappropriate for the more deep-going "Principle-Theories": "By and by I
despaired of the possibility of discovering the true laws by means of con-
Einstein's Model for Constructing a Scientific Theory 115

structive efforts based on known facts. The longer and more despairingly I
tried, the more I came to the conviction that only the discovery of a univer-
sal formal principle could lead us to assured results" (Schilpp, p. 53).4

5 Two logical discontinuities in J

We have to linger a little more on the implications of the trajectory,

are, ascent, or jump (to which we will now assign a label, J). As Einstein
often stressed (cf. analysis of Autobiographical Notes, and many other
sources, e.g., l.0., p. 291), there are in fact two sets of logical discontinuities
implied in the seemingly smooth curved line. We fashion it by fastening our
attention on "certain repeatedly occurring complexes of sense impressions"
(1.0., p. 291) and "relating to them [zuordnenJ a concept" (ibid.). The con-
cept is then a kind of "mental knot" or "mental connection" (ibid.) bet-
ween sense impressions, and is "primary" (1.0., p. 293) if close to sense
experience. But we select the concept without some logical necessity, really
"arbitrarily" (I.O., p. 291) in the sense that "considered logically this con-
cept is not identical with the totality of sense impressions referred to; but
it is a free creation of the human (or animal) mind" (ibid. - "human or
animal mind": another unnecessary barrier unceremoniously discarded!).
The same theme of the logical discontinuity in the formation of con-
cepts appears again and again. For example: "All concepts, even those
which are closest to experience, are from the point of view of logic freely
chosen conventions" (Schilpp, p. 13). And again: "There is no inductive
mf'thod which could lead to the fundamental concepts of physics. Failure
to understand this fact constituted the basic philosophical error of so many
investigators of the nineteenth century" (1.0., p. 307). Several times Ein-
stein referred to David Humes attack on induction, showing that "concepts
which we must regard as essential, such as, for example, causal connection,
cannot be gained from material given to us by the senses" (1.0., p. 21).

4 The schema in the letter to Solovine applies to fashioning Principle Theories. That
does not mean Einstein never fashioned Constructive Theories. An example is, in
fact, the photon theory, including the explanation of the photoelectric effect, which
he had published earlier in 1905. But precisely because he did not view the introduc-
tion of the quantum as anything more fundamental than a "heuristic point of view"
to deal with the problem at hand, he could never bring himself to regard quantum
physics as more than a transient analysis. Only in Theories of Principle, he thought,
are the postulated axioms far enough above the plane of experience, and far enough
from ad hoc reasoning, to yield a creative scheme that can handle the totality of
the facts of experience On Einstein's view concerning Principle Theories versus
Constructive Theories, which he discussed from 1919 on, see 1.0., pp. 228, 302-303,
318-319, and Holton, Thematic Origins. pp. 230,252,316 and 348.
116 Gerald Holton

To the same end, Einstein also reminded his readers frequently (e.g.,
1.0., p. 298) of the fatal error that had been made for so long in thinking
that the basis of Euclidean geometry was logically necessary; this error was
caused by forgetting the empirical base and hence the limited experiential
context within which all concepts are fashioned. A similar illusion was the
great obstacle to formulating the Special Relativity Theory (1.0., pp. 298-
299), namely that there exists a universal time applicable to all events in
space as a whole, a concept of time long held to be an a priori given, neces-
sary conception, seemingly independent from our sense experience. This
error was caused by forgetting that the r.otion of time itself arises initially
in our everyday experience by watching sequences of events happening at
one locality, rather than in all of space.
Deprived of any certainty that our concepts have a necessary connec-
tion with the corresponding experiences, we begin to see the precariousness
of the business of theory construction. But we can do no better. We create
new concepts, perhaps suggested at first only tentatively, and gather them
together with old concepts whose usefulness has been tested in previous
struggles, knowing that neither one nor the other is sacred and unchangeable,
neither induced nor in any other way securely abstracted from the plane of
experiences below. It may be that this discontinuity is symbolized by the
small gap in the drawing between the horizontal line E and the arc rising from
that region to A above.
There is a second logical discontinuity which also enters to make it
a "mistake to permit theoretical description to be directly dependent upon
acts of empirical assertions" (Schilpp, p. 674). This concerns the relation
of concepts to one another when they are used together to make a system
of axioms - for example, some postulated laws of nature ("propositions
expressing a relationship among primary concepts" 1.0., p. 293). Not only
each individual concept, but the whole "system of concepts is a creation
of man" (Schilpp, p. 13), achieved in a "free play," the justification for
which lies only in the pragmatic success of the scheme being built up to
give ultimately a "measure of survey over the experience of the senses
which we are able to achieve with its aid" (Schilpp, p. 7).
The two-fold discontinuity, then, is a good part of the reason why
Einstein repeats, again and again, sentences like this one from 1918: "There
is no logical path to these elementary laws; only intuition, supported by
being sympathetically in touch with experience [Einfiihlung in die Erfah-
rungj" (1.0., p. 226). The repeated insistence was in good part in opposition
to the then current form of positivism, which, for example, saw the goal
of scientific work to be the economic statement of relations among obser-
vables. To this day, Einstein's formulation causes hostility from some
philosophical quarters, which insist on exaggerating this particular element
in Einstein's total schema. (On the other hand, it should also be said that
Einstein's Model for Constructing a Scientific Theory 117

Einstein's anti-inductivism has encouraged some of the most interesting

contributors to philosophy of science in our day.)
In no way can Einstein's message on this point be taken to celebrate
irrationality, to give primacy to intuition, or the like. Rather, it represents
two truths which he had learned, so to speak, on his very own body. One
was the liberating warning that just because all theories are "manmade"
and "the result of an extremely laborious process of adaptation," they
are also "hypothetical, never completely final, always subject to que-
stion and doubt" (1.0., p. 323). The other message was, precisely against
this somber knowledge, to encourage the assertion of ingenuity and inno-
vation, in science as well as outside, if necessary against prevailing dogma.
(Einstein quipped: "Would Faraday have discovered the law of electro-
magnetic induction if he had received a regular college education?" 1.0.,
p. 344.) If accused of dragging down, from the Olympian fields, "the fun-
damental ideas of thought in natural science, and to attempt to reveal
their earthly lineage" (1.0., p. 365), Einstein would answer that he did so
"in order to free these ideas from the taboo attached to them, and thus
to achieve greater freedom in the formation of ideas and concepts. It is
to the immortal credit of D. Hume and E. Mach that they, above all others,
introduced this critical conception" (ibid.).s

6 Constraints and freedoms

We might point to other properties of the concepts by means of which

axioms are formulated. Although Einstein does not stress it often explicitly,
one type of conceptual construction needed to keep the concept from float-
ing away like some arbitrary soap bubble is the definition which we give to
every abstract term (point, length, time interval, electric charge). While the
definition of any term is logically arbitrary, it is connected with observables
by our "operational definition" or "semantical rule" to which we agree to
adhere once it has been fashioned. As early as 1916, Einstein wrote "Con-
cepts have meaning only if we can point to objects to which they refer and
to the rules by which they are assigned to these objects". I
Good examples of this operational approach to concepts can be found
in Einstein's own careful analysis of the mental (mathematical) and physical
operations of measuring time in his first relativity paper, or in his description
of what is meant by such conceptions as "solid body" or "space" in several

5 However, Mach, with W Ostwald, is scolded by Einstein (Schilpp, p.49) for his
"positivistic philosophical attitude" which misled them into opposing atomic theory.
They were victims of "philosophical prejudice," chiefly "the faith that facts by
themselves can and should yield scientific knowledge without free conceptual con-
struction. "
118 Gerald Holton

of his later essays. Therefore, one might elaborate the diagram in Fig. 1 by
drawing thin, vertical lines between E and A, to indicate that such connec-
tions are made whenever we choose the convention or "meaning" assigned
to a term that is part of the scientific (or any) vocabulary.
The other constraint on our choice of concepts - even though they
"have a purely fictious character," being the "free inventions of the human
intellect, which cannot be justified either by the nature of that intellect or
in another fashion a priori" (1.0., p. 272) - lies in Einstein's call for fruga-
lity and simplicity. After all, the aim of any good theoretical system is "the
greatest possible sparsity of the logically independent elements (basic con-
cepts and axioms)" (Schilpp, p. 13). Any redundancy or elaboration must
be avoided, for "it is the grand object of all theory to make these irreducible
elements as simple and as few in number as possible" (1.0., p. 273). For
example, it was, in his view, "an unsatisfactory feature of classical mechanics
that in its fundamental laws the same mass appears in two different roles,
namely as inertial mass in the laws of motion, and as gravitational mass in
the law of gravitation" (1.0., p. 308). The equivalence of these two inter-
pretations of mass signaled to him a truth which needed to be stated as a
basic axiom (in General Relativity Theory), rather than saddling the theory
with a proliferation which did not seem to be inherent in the phenomena.
In good part as a result of Einstein's own work and example, physical
scientists have indeed succeeded in showing, during this century, that only
a very small number of postulated fundamental laws, employing a sur-
prisingly small number of fundamental concepts, are needed to encompass
("explain") at least in principle an ever-growing infinity of separate facts of
experience. This does not mean at all that everything is explained, or even in
principle already explainable; but still, it is a "wonder," and a motivation
for further work. This success also has some curious consequences, to which
we shall return later.
One corrolary of this method of hypothesizing is that during the period
of constructing a theory the innovator must give his proposed "jump" to the
axioms a chance to prove itself. Hence in this early and usually private stage
of theorizing the researcher has to grant himself a freedom, the right of "sus-
pension of disbelief," a moratorium on premature attempts at falsification
(i. e., on attempts to discredit the hypothesized postulate by disproving it). 6

6 For a discussion of the concept of "suspension of disbelief," see Holton, The Scien-
tific Imagination: Case Studies (New York and Cambridge: Cambridge University
Press. 1978), pp.71-72. Even the "dean" of logical positivists of his day, Hans
Reichenbach, might have agreed, for he said "The physicist who is looking for new
discoveries must not be too critical; in the initial stages he is dependent on guess-
ing, and he will find his way only if he is carried along by a certain faith which
serves as a directive for his guesses," etc. (Schilpp, p. 292). But he went on to deny
that such mechanisms can or should be of interest to "the philosopher of science."
Einstein's Model for Constructing a Scientific Theory 119

A related corrolary is more disturbing, however. Just as there were in

principle infinitely many points on the E level at the bottom of the schema,
there are in principle infinitely many possible axioms or systems of axioms
A at the top. The choice a given scientist makes out of all possibilities
cannot be entirely arbitrary, since it would easily involve him in an infinitely
long search. How does one in fact make this choice? That is, what guides
or constraints do exist which help (or hinder) the innovator in making his
particular jump to A rather than to some other A I, which another researcher,
on the basis of the same E. may prefer to make? Einstein says nothing about
it in this letter; but he wrote sufficiently elsewhere to help us deal with the
question, as we shall see later.

7 The logical path

To return now to Einstein's letter to Solovine. He continues his expla-

nation of the schema: "3. From A, by a logical path, particular assertions
are deduced- deductions which may lay claim to being right."
This sentence takes us to the area in the schema where rigorous analy-
tical thinking enters. where the scientific imagination indeed requires the
"logical path." "Logical thinking is necessarily deductive" (La., p. 307),
starting from the hypothetical concepts and axioms which were postulated
during the earlier upward swing of the schema. Therefore we are now pro-
ceeding downward from the axioms, deriving the necessary consequences
or predictions; if A. then 5, 5', 5" ... should follow; or as in the 1905
relativity paper, if the Principle of Relativity and the Principle of the Con-
stancy of the Velocity of Light are assumed in the first place as axiom
system A, then there follow necessarily, without any further fundamental
assumptions, the transformation equations for space and time coordinates,
the relativity of simultaneity, the so-called length contraction and time
dilation effects, and, at the end of the paper, "the properties of the motion
of the electron which result from the system of equations and are accessible
to experiment .... tThese I relations are a complete expression for the laws
according to which. by the theory here advanced, the electron must move."
This wealth of results is a natural consequence of Einstein's powerful deduc-
tions. In an excellent paper entitled "Logical Economy in Einstein's 'On
the Electrodynamics of Moving Bodies.'" Robert B. Williamson 7 has clearly
shown the logical consistency and parsimony of Einstein's detailed argu-
ment. These features make it even more plausible that the whole work
represents the crystallization of years of effort.

7 Stud. Hist. Phil. Sci. 8 (I 977), pp. 49-60. See also P. Mittlestaedt, "Conventionalism
in Special Relativity." Foundations of Physics. 7 (1977), pp. 573-583.
120 Gerald Holton

Some who have criticized Einstein's remarks as giving too much weight
to intuition and other logically speculative concepts have tended to overlook
the definite role which Einstein did give to the logical phase of the scientific
imagination. If he argues for recognizing the necessary inspirational compo-
nent in the formation of fundamental hypotheses at the level of A, he also
goes on to say "the structure of the system is the work of reason" (1.0.,
p. 272). This part of the scientist's work, where inference follows inference,
requires "much intense, hard thinking," (1.0., p. 282) but at least is a task
that one can learn in principle "at school" (1.0., p. 221). It is only the earlier
step, that of establishing the principles in the first place from which deduc-
tion can proceed, for which "there is no method capable of being learned
and systematically applied .... the scientist has to worm these general prin-
ciples out of nature" (ibid.).

8 Testing against experience

Continuing in his letter to Solovine, Einstein now comes to the fourth

and final step that brings us back to the plane from which we started:
"4. The 5 are referred [or related] to the E (testing against experience)."
Still anxious to make the necessary distinction between what logic
can and cannot be expected to do during the process of theory construction,
Einstein adds parenthetically: "Carefully considered, this procedure also
belongs to the extra-logical (intuitive) sphere, because the relations between
concepts appearing in 5 and the experiences E are not of a logical nature.
[Perhaps this is why Einstein draws the vertical arrow-tipped lines from
5, Sf, ... as dotted lines.] This relation of the 5 to the E is however (prag-
matically) far less uncertain than the relation of the A to the E. (For exam-
ple, the concept 'dog' and the corresponding experiences.) If such corres-
pondence could not be attained with great certainty (even if not logically
graspable) the logical machinery for the 'comprehensibility of reality' would
be completely worthless (example, theology).
"The quintessence is the eternally problematic connection between the
world of ideas and that of experience (sense-experiences)."
The main point of interest in this passage is the first sentence, "The
5 are referred to the E." Even its simplicity of expression does not hide the
difficulties in the content. We are now at the crucial last phase of the sche-
ma, and we are looking down, from the predictions and other consequences
(5, Sf ... ) of the partly hypothesized, partly deduced scheme, to find whether
corresponding observations can in fact be found to exist on the plane of ex-
perience E. If these are found, we can say that our various predictions have
been borne out by observation and that we therefore have a right to regard
with more confidence the previous steps that led us to this last one - the
jump J from E to A, the postulation of A, and the deduction of the 5. We
Einstein's Model for Constructing a Scientific Theory 121

thus have completed the cycle implied in the schema E -+ j -+ A -+ 5 -+ E.

For simplicity, I will refer to this schema as Einstein's EjASE process of
scientific theory construction.
But Einstein knew well that even if the predictions are borne out,
one must not be too confident that the theory, the whole structure of
conjectures, postulations and deductions, is necessarily right. This is so for
three reasons. First, right predictions can be drawn from wrong axioms.
Hence theories that have turned out to be fundamentally in error (Aristo-
telian theory of elements, phlogiston theory, caloric theory) nevertheless
were for a long time thought to be to be "verified" by the coincidence of
deductions and observation.
Second, it is even impossible in principle to consider a theory "proven"
once and for all, since this would entail subjecting it to an infinity of tests
by observation, and not just now but for all future times. There is no such
thing as a final verification or confirmation of a theory by experiment or
observation. The most one can ever claim is that a theory gains more and more
plausibility or usefulness the longer the various predictions derivable from
it are found to correspond to the growing area of available sense experience
- and the fewer the contradictions.
Third, and most important, Einstein came to realize that except per-
haps in the simplest cases, one cannot rely on what someone claims to be
"experimental facts" without much probing. The "confirmations" of theo-
ries have often turned out to be the result of misinterpretation of the data
or a malfunctioning of the experimental apparatus. Einstein had more
than once been held up in this theoretical work by claims of experimental
scientists that turned out to be wrong. As he was reported to have said in
the mid-1920s:
"You must appreciate that observation is a very complicated process.
The phenomenon under observations produces certain events in our measur-
ing apparatus. As a result, further processes take place in the apparatus,
which eventually and by complicated paths produce sense impressions and
help us to fix the effects in our consciousness. Along this whole path -
from the phenomenon to its fixation in our consciousness - we must be
able to tell how nature functions, know the natural laws at least in practical
terms, before we can claim to have observed anything at all."8

9 Criteria for a good theory: I. "External Validation"

What then can one expect to be the proper relation between the 5
and the E in an adequate theory, at least one of the more ambitious kind
that was of interest to Einstein, a theory whose object is the "totality

8 W. Heisenberg, Physics and Beyond (New York: Harper and Row, 1971), p. 63.
122 Gerald Holton

of physical phenomena" (Schilpp, p. 22)? As we saw, in Einstein's sen-

tence "The S is referred to the E," the phrase "referred to" [in Beziehung
gebracht] is not at all the same thing as "verified against," as one might
expect to read here if the proposed test of a good theory were the test of
"verification." But that would be a pre-twentieth century view which has
turned out to be too optimistic an estimate of the solidity of scientific
theories.
In fact, Einstein had proposed, some years earlier, two criteria for
a good theory, two tests "according to which it is at all possible to subject
a physical theory to a critique" (Schilpp, p. 20). The first test is what
Einstein called the criterion of "external validation," and it is "concerned
with the validation [Bewa'hrung] of the theoretical foundations by means
of the material of experience [Erfahrungsmaterial] lying at hand" (Schilpp,
p. 22).9 The criterion is simply this: "The theory must not contradict
empirical fact" (Schilpp, p. 21).
Note that this is a principle of disconfirmation or of falsification, and
hence much more sophisticated than any injunction to seek "confirma-
tion" by empirical test. It is more generous, because in the absence of
disconfirmation one can hold on to the theory - "Once a theoretical idea
has been acquired, one does well to hold fast to it until it leads to an un-
tenable conclusion" (1.0., p. 343) - and it is also a sharper demarcation
criterion because the presence of believable disconfirmations soon dis-
credits a theory, whereas a continued absence of verification merely delays
the final decision.
The disconfirmation criterion does not mean at all, however, that
presumed confirmations, or coincidences of S with corresponding elements
of E, would be unwelcome. On the contrary; in fact, the purpose of the
majority of actual experimental investigations is guideq by the hope of
finding correspondence of this type by which the plausibility of some
previously examined theory would be increased. But for the reasons given
above, absence of verification is not conclusive either way, permitting
one to be skeptical about or holding on to the theory until further notice,
depending on one's prejudice. But what really decides the matter is stubborn
and repeated evidence of disconfirmation.
And it really should be stubborn and repeated evidence. One cannot
abandon a theory every time a disconfirmation is reported. That would
be extreme experimenticism, not warranted by the delicate and difficult
nature of experiments in modern science. One should be as reasonably

9 Wherever possible I have checked the published English-language translation of

material Einstein published first in German, and where necessary have corrected
the translation, as in this case.
Einstein's Model for Constructing a Scientific Theory 123

skeptical about experiments that disconfirm as about those that confirm -

and particularly if the experimental disconfirmation of one theory is used
to support another which, on other grounds, is less appealing.

10 Criteria for a good theory: II. "Inner prefection"

What can this other ground be? What can make a theory more appealing
or less so, other than the criterion of "external validation"? The answer
is given by Einstein's second criterion for subjecting a theory to a critique.
He called it the criterion of "inner perfection," and it concerns itself with
choosing the superstructure in the EjASE scheme, namely, the j, A, and
S. One must remember that there is no guarantee in a given case that the
elements of a theory are unique. It often happens that two quite diffe-
rent theories, with different J. A, and S, arise out of concern with the same
material of experience, and moreover give equally good correspondence
between their sets of S and the relevant sense experiences. The most famous
case is of course that of Ptolemaic theory and the theory of Copernicus in
the 16th Century. While different with respect to their basic axioms, both
theories arose out of the need to account for the same regularities and ir-
regularities in E, in the observed motion of celestial bodies, and the predic-
tions derivable from both theories had about the same degree of correspon-
dence with the observables.
Einstein's second criterion was frankly stated in his Autobiographical
Notes. "The second point of view is not concerned with the relation to
the material of observation but with the premises of the theory itself,
with what may briefly but vaguely be characterized as the 'naturalness'
or 'logical simplicity' of the premises (of the basic concepts and of the
relations between these which are taken as a basis)" (Schilpp, p. 23).
This of course is not an entirely new idea; Einstein acknowledges
that it "has played an important role in the selection and evaluations of
theories since time immemorial." But in practice, the requirement of natu-
ralness or logical simplicity, or "unity and parsimony" (1.0., p. 23) has
never been easy to follow. Einstein is warning here to stay clear of theories
that are patched up by ad hoc assumptions introduced just to make the de-
ductions correspond better to the facts of experience as they continue to
come in. "For it is often, perhaps even always, possible to adhere to a
general theoretical foundation by securing the adaptation of the theory
to the facts by means of artificial additional assumptions" (Schilpp, pp. 21-
23). Early in his career, Einstein had considered Lorentz's theory of the
electron as just such a patchwork in the sense that it avoided factual discon-
firmation only by introducing assumptions specially chosen for this very
purpose (introduction of length contraction to explain absence of predicted
effect in the ether-drift experiment). This practice could be represented by
124 Gerald Holton

A1

JC~s,- ,
*
S' SOl
, I
I

E * E1
t
E2
E----~---±--------+_
* E1 *E2 E3
Schema C1 Schema C2

Figure 2

a modification in the diagram of the ElASE process, as in Fig. 2, where

schema C1 is modified to yield schema C2 by changing A to (A + ~), ~
being the modification in A introduced to deal with the problem of obtain-
ing better correspondence between the 5ieductions S and the facts E.
To be sure, theories are likely to grow in some such manner, as they
are applied to new areas of phenomena. And in any case such criteria as
"naturalness" or "logical simplicity" or "economy" or "unity and parsi-
mony" are not easy to defend or even specify, for their "exact formula-
tion ... meets with great difficulties" (Schilpp, p. 23).
It requires from us not a mere "enumeration of logically independent
premises," but "a kind of reciprocal weighing of incommensurable qualities,"
(ibid.) hence a judgement into which esthetic considerations and other pre-
ferences can enter prominently.
Einstein was aware of a paradox, in that he tried to deal with great and
complex areas of varied experience, and yet looked "for simplicity and
economy in the basic assumptions. The belief that these two objectives can
exist side by side is, in view of the primitive state of our scientific know-
ledge, a matter of faith .... This, in a sense religious, attitude of a man
engaged in scientific work has some influence upon his whole personality"
(1.0., p. 357). Again, writing at about the same time (1950) elsewhere,
he acknowledged the a priori implausibility "that the totality of all sensory
experience can be 'comprehended' on the basis of a conceptual system built
on premises of great simplicity. The sceptic will say that this is 'a miracle
creed.' Admittedly so, but it is a miracle creed which has been borne out to
an amazing extent by the development of science" (1.0., p. 342).
An example of Einstein's commitment to simplicity and naturalness
among the fundamental conceptions of science - an example that haunted
Einstein for much of his scientific life - was his unshakable dislike for the
premises and program of quantum mechanics. For the mathematical des-
cription in quantum mechanics deals in principle with statistical ideas (e.g.,
densities in the ensemble of systems), eliminating thereby even the possi-
bility in principle of describing the detailed behavior of the single object
or system - the very thing that lies close~t to our experience, as in the
Einstein's Model for Constructing a Scientific Theory 125

sensory reports made available by cloud chambers, counters, and the like.
To believe that this program is right "is logically possible without contra-
diction; but it is so very contrary to my scientific instinct that I cannot
forego the search for a more complete conception" (1.0., p. 318; and also
often similarly, e.g., 1.0., p. 316; letters to Max Born, discussion with
Niels Bohr, etc.). This search for a more complete conception, he knew,
might be doomed. "In the end the choice will be made [by the profession
as a whole 1 according to which kind of description yields the formulation
of the simplest foundation, logically speaking." But until the evidence is
irresistible, he considered it his right to abstain from "the view that events
in nature are analogous to a game of chance. It is open to every man to
choose the direction of his striving" (1.0., pp. 334-335).
Einstein's use of the colorful words "instinct," "striving," "intuition"
or "wonder" was not intended as a calculated provocation of some scien-
tists or philosophers, but is has had this effect nevertheless. To make matters
worse, he referred to yet another process of importance in the growth of
theories, known to every practicing scientist but difficult to define. For
even though he acknowledged that these two criteria of "external valida-
tion" and "internal perfection" defied precise description, he held that
among the "augurs," those who deal deeply with physical theory, there
nevertheless exists at any given time agreement in judging the degree of
external validation and inner perfection (cf. Schilpp, pp.23-25). Once
more, the absence of an air-tight definition did not preclude him from
putting his bet on the usefulness of a concept, in this case that of consensus
in groups within the scientific community.

11 Going beyond the precis

In putting together this account of Einstein's own epistemological

views, insofar as possible in his own words, I have been trying to do justice
to his meticulous sense of the realities, the lack of guarantees, the tenta-
tive, fallible, human aspects of each element in theory construction, and
of the "eternal antithesis in our area between the two inseparable compo-
nents of our knowledge, the empirical and the rational" (1.0., p. 271). The
schema that has emerged is miles from the self-confident and axiomatic
treatments of scientific methodology which Einstein rightly held to have
little resemblance to the actual practice of the working scientist. But let us
not make a mistake in the opposite direction. Despite all its disclaimers,
Einstein's schema implies nothing less than a description for as solid a
process of reasoning as is in fact available to scientists.
Of course, Einstein's letter to Solovine was not a solemn publication
but a precis shared between friends. But it is tremendously suggestive, and
invites, in the spirit of Einstein's own method, to go further beyond these
126 Gerald Holton

ideas, and so to see how they can serve to handle other problems of theory
construction and the scientific imagination. As in science itself, our belief in
a scheme increases if we find that it is not merely ad hoc for covering the
area within which it was specifically proposed but is successful beyond it.
There are specifically two problems that Einstein's schema can help us with:
how scientific theories grow and give way to other theories, and how to
understand better the controversies involving fundamentally different
theories that claim to deal with the same experimental facts.

12 The growth of a theory

We noted that the schema in Fig. 1 is not a static one, but is a process
which makes a cycle from E via j, A, 5, back to E. But a theory can hardly
be created and tested by going through the cycle once. Even the theories
by which we orient ourselves in our day-to-day life, and a fortiori the estab-
lished theories of science which we honor and use as tools that have come
down to us from earlier workers and controversies, are all the results of
cycles of progressive adaptation, making them more acceptable by using
the feedback from one cycle to modify the next. Moreover, this process
of modification and growth will continue as new phenomena are found that
enlarge the original area of application. Physics is constantly "in a state of
evolution .... Evolution is proceeding in the direction of increasing simpli-
city of the logical basis" (1.0., p. 322).
The need to go through many cycles (C 1 -+ C2 -+ C3 ••• ) of the EjA5E
process is forced on us, if by nothing else, by our human limitations. Neither
thought by itself nor sensory experience by itself leads to reliable human
knowledge. For concepts can be subjected to analysis which gives us cer-
tainty of the kind "by which we are so much impressed in mathematics;
but this certainty is purchased at the price of emptiness of content" (1.0.
pp. 276-277). On the other hand, sense experience cannot be related to the
concepts, as we have seen, except by adopting essentially arbitrary defini-
tions (conventions), and hence they cannot claim certainty either. The best
we can therefore do is to let whatever trustworthiness there is in our theory
construction come out of the interplay of thought and sense experience
through many cycles, carried out over time. Theories therefore have to be
"thoroughly elaborated" (1.0., p. 282) and have to evolve - first in the mind
of the innovator before publication and then in the community of scientists
through discussion or controversy.
For example, in going through the first cycle of the schema, the 5 of
the theory at that stage (SI, S~, S~' ... ) may show an incomplete correlation
with the "facts" in the E plane. Einstein gives an example which, he says,
is one of the considerations which "kept me busy from 1907 to 1911": in
his early attempts to generalize relativity theory, "the acceleration of a
Einstein's Model for Constructing a Scientific Theory 127

falling body was not independent of its horizontal velocity or the internal
energy of the system," contrary to the "old experimental fact" (1.0., p. 287).
A discrepancy of this sort forces one to rethink the A, modifying the original
axiom system At to become a somewhat different one,A 2 •
We recall that Einstein warned that such modification should be made
not in a merely ad hoc. burte-force way, but for example by recasting the
axiom system into a more generalized form that permits more deductions
S2, S;, S;' ... that can be correlated with the E, and if possible from fewer
independent concepts. Thus Einstein was able to go from the first principle
of restricted relativity theory, that all natural laws must be so conditioned
that they are covariant with respect to Lorentz's transformations, to the
first principle of general relativity theory, that natural laws are to be for-
mulated in such a way that their form is identical for coordinate systems
of any kind of states of motion. (1.0., pp. 329~330). In this way Einstein
removed his dissatisfactions with the special nature of the original relativity
theory, namely that it referred only to systems in uniform motion to which
no absolute significance could be attached. The introduction of the principle
of equivalence removed the contradiction between the predicted acceleration
of a falling body and the observed one, as well as removing an unneccessary
duplication (two meanings of mass, as referred to above).
Fig. 3 represents schematically the progress from the early state of
a theory to a later state, from C1 to C2 and from C2 to C3 • Here C3 could
stand for the next step which Einstein saw needed after his success in fash-
ioning general relativity theory; for he felt that "the theory could not rest
permanently satisfied with this success .... The idea that there exist two
structures of space independent of each other, the metric-gravitational and
the electromagnetic, was intolerable to the theoretical spirit" (1.0., p. 285).
Hence, Einstein's persistent attempt to fashion a field theory that corres-
ponds to a "unified structure of space" (ibid.). Again and again, the word
"unity" beckons as Einstein's final goal ~ "seeking, as far as possible, logical
unity in the world picture, i.e., paucity in logical elements" (1.0., p. 293);
"thus the story goes on until we have arrived at a system of the greatest
conceivable unity, and of the greatest poverty of concepts of the logical
foundations, which is still compatible with the observations made by our
senses" (1.0., p. 294).

, s'2"
E-----+-----4----~~----~
E, E4
t
Schema C,
Figure 3
128 Gerald Holton

To give some other examples of the driving force leading from C 1

to C2 to C3 : Einstein held Newton's mechanics to be "deficient" from
the point to view of the requirement of the greatest logical simplicity of
the foundation insofar as the choice of the value 2 of the exponent in the
inverse square law of gravitation - the very heart of Newton's greatest
triumph - was heuristic or ad hoc in the sense that it could be defended
only because it worked. And in addition the law of gravitation itself was
a separate postulate, not connected with or derivable from other concep-
tions in mechanics - whereas in general relativity theory it developed
as a consequence of the postulates. Similarly, Einstein felt that H. A. Lo-
rentz's synthesis of Newtonian mechanics and Maxwell's field theory con-
tained the "obviously unnatural" mixture of total differential equations
(for the equations of motion of particles or points) with partial differen-
tial equations (Maxwell's field equations). It led to the need to assume
particles of finite dimension, to keep fields at the surface from becoming
infinitely large. To Einstein it appeared "certain ... that in the foundations
of any consistent field theory the particle concept must not appear in addi-
tion to the field concept" (1.0., p. 306).
One could also visualize the progressive development of scientific
theory to take place as the development of the system of concepts at an
increasingly higher level of "layers" or strata, each layer having less and
less direct connections with the complexes of sense experiences (1.0.,
pp. 293-295). In this way, a more phenomenological theory at the early
stage of science, e.g., the theory of heat before Maxwell, gives way to a
more independent set of concepts and axioms that characterizes, for example,
kinetic theory and statistical mechanics. Thus the latter eventually allowed,
in the study of Brownian motion, to find the limits of applications of the
laws of classical theory, and in addition provided a definite value for the
size of atoms and molecules, obtained by several independent methods.
There is of course a cost in this developmental process. By going cycli-
cally through several stages of theories, each stage is forced to use conceptions
more removed from direct experience (e.g., atomism). The distance from the
E to the A is larger, the contact with common sense is more and more tenu-
ous. But the fundamental ideas and laws of science attain a more and more
unitary character (cf., 1.0., p. 303). Eventually all the sciences would attain
this final stage.
In the meantime another cost of this process is that the more general
the theory becomes the longer it may have to wait for the correlation of
its predictions with the ground of experience. Thus it took the general
theory of relativity to 1919 to make proper contact with E. The delay
may well test the self-confidence of the theoretician to the utmost. "It
may need many years of empirical research to ascertain whether the theo-
retical principles correspond with reality" (1.0., p. 222); for it may take
that long to discover the "necessary array of facts" (1.0., p. 223).
Einstein's Model for Constructing a Scientific Theory 129

13 Representations of a finished theory

In the ordinary course of events the development of a theory will take

it to a stable canonical form. It enters the textbooks usually as a recast
pedagogic scheme which is characterized by a rearrangement to bring out
an axiomatic structure and to hide all traces of the speculative phase that
motivated and characterized the theory in its early stage. In particular,
the textbook tends to hide the] process as if it were an embarrassment.
The presentation of the theory at that stage in its life cycle, and of scien-
tific research papers that base themselves on such a theory,IO and likely
to look like Fig. 4. That is, a few phenomena are cited (E I , E2 in Fig. 4)
from which, it is said, the axiom system was induced; and from the latter,
predictions are deduced for which corresponding experimental demon-
strations (E 3, E4 ... ) can in fact be given. Or, as in Fig. 5, the whole theory
is presented as if its starting point is the discovery of an axiom system,
from which all follows. That is essentially the style of the Principia of
Newton and of most school books; e.g., Newton's laws of motion and of
gravitation, at the top, radiate deductions concerning the periodicity of
the tides, the shape of planets, etc., and these in turn are directly confirmed
by the experimental evidence below. Or from the fundamental postulates
of the kinetic theory, there follow the equations of state of gases, viscosity,
diffusion, heat conductivity, radiometric phenomena in gases, etc; and sure
enough, all these can be correlated with their corresponding phenomena
over a large range.
Of course, apart from the fact that such a representation no longer
reflects the genesis of the theory (which, in any case, is usually of as little
interest to most scientists as to most philosophers) a representation as that
in Fig. 5 brings out the astounding strength of well-developed theories.
A A

5
5" I I I I
I I I I I
E E t t I t
El E2 E3 E4 E5 *
Figure 4 Figure 5

10 The commo nly agreed-upon structure of writing scientific papers for publication,
which makes it seem that the gathering of data and induction from them formed
the beginning of scientific work, has prompted P. B. Medawar to call the scien-
tific paper a "fraud" and a "travesty of the nature of scientific thought." P. B.
Medawar, "Is the Scientific Paper a Fraud?," The Listener (1963), pp. 377-378.
130 Gerald Holton

That is, they give us, to use Einstein's phrase, an "overview" by which the
multiplicity of immediate sense experiences of the most diverse kinds are
brought into a unified and therefore understandable scheme.

14 Progress by unification of theories

The next stage in the historic progress of science occurs when a unifi-
cation of two or more theory systems is forged, as when Galileo joined
terrestrial and celestial physics, or when Maxwell produced a synthesis
of electricity, magnetism and optics. Before unification or synthesis, each
of the theory systems has its own system of concepts, and while they might
be closer to experience than the concepts after unification, they lack unity
among their different fundamental postulates within which they are em-
bedded (cf. 1.0., p. 302). In terms of our graphic shorthand, what happens
is symbolized by Fig. 6. On the left side, the separate axiom systems for the
fields of electricity, magnetism,. and optics dominate the respective systems
that sit, like separate pyramids, upon their corresponding territories on the
E plane. After Maxwell's synthesis, the separate axiom systems become
merely special cases of a more general one, incorporating Maxwell's equa-
tions. The separations between the three areas on the plane of phenomena
disappear, the sets of facts previously connected become part of one larger
set, and their theoretical description has thereby been simplified (cf., 1.0.,
p.223).
The same symbolization would be appropriate for many of the great
advances of science, for example, the work of P. A. M. Dirac in the late
1920s by which large areas of both physics and chemistry were brought
under the control of quantum mechanics. In our time, the current attempts
to unify the basic forces of nature is another chapter in this drive to en-
compass eventually the totality of E, all points on the E plane, in terms
of the fewest possible independent axiom systems. It was of course Einstein's

Figure 6
Einstein's Model for Constructing a Scientific Theory 131

hope through field theory to establish a unified foundation for the whole
of physics (cf., 1.0., pp. 328-329). Einstein put this aim very clearly: "From
the very beginning there has always been present the attempt to find a
unifying theoretical basis for all these single sciences, consisting of a mini-
mum of concepts and fundamental relationships, from which all the con-
cepts and relationships of the single disciplines might be derived by [aJ
logical process. This is what we mean by the search for a foundation of
the whole of physics. The confident belief that this ultimate goal might be
reached is the chief source of the passionate devotion which has always
animated the researcher" (1.0., p. 324). It may well be that the ability to
carry in mind the entities and relations symbolized in the right-hand portion
of Fig. 6 is what it means to "have a Weltbild," to have a mental picture of
the physical universe.

15 The role of thematic presuppositions

We now must come back to deal with an important point which was
left open. The problem can be put as follows: Since the leap from E to A
at the beginning of the schema in Fig. 1 is logically discontinuous and
represents the "free play" of imagination, and since from such a leap can
result an infinite number of A - virtually all of which will turn out to
be useless for the construction of the theory system - how can one ever
expect to be successful in this process except by chance? The answer must
be that the license implied in the J process is the freedom to make a leap,
not the freedom to make any leap at random. Something must guide or
channel J if only because the premises later must pass such tests as those
of naturalness and simplicity in order to meet Einstein's second criterion
for a good theory.
The chief guide is a constraint that shapes the work of every scientist
engaged in a major work on novel ground: the constraint provided by
explicit or, more usually, implicit preferences, preconceptions, presuppo-
sitions. Einstein himself recognized and commented on this repeatedly:
"If the researcher went about his work without any preconceived opinion,
how should he be able at all to select out those facts from the immense
abundance of the most complex experience, and just those which are simple
enough to permit lawful connections to become evident?"l1 By way of
example, he discussed the dilemma that, in formulating the laws of mecha-
nics, one has to follow either the "natural tendency of mechanics to as-
sume .. , material points," which necessarily leads to the presupposition of

11 A. Einstein, "Induktion und Deduktion in der Physik," Berliner Tageblatt. 25 De-

cember 1919.
132 Gerald Holton

atomism, or else to erect a mechanics of continuous media based on an-

other "fiction," e.g., that "the density and the velocity of matter depend
continuously upon the coordinates and time" (1.0., p. 302). These "fic-
tions" - not unrelated to what Frank Kermode in another context has
called the "necessary fictions" that are found at the heart of literary works
- have of course considerable practical value. For example, they guide
the development of mathematical tools (in Einstein's last example, partial
differential equations), but they are much more than that. He referred
to them as " 'categories' or schemes of thought, the selection of which is,
in principle, entirely open to us and whose qualification can only be judged
by the degree to which its use contributes to making the totality of the
content of consciousness 'intelligible' "(Schilpp, p. 673).
An example of such a category is the distinction between sense impres-
sions and "mere ideas" (ibid.). He warned that "we do not conceive of the
'categories' as unalterable (conditioned by the nature of the understanding
[and in this respect "distinct from that of Kant"]) but as (in the logical
sense) free conventions. They appear to be a priori only insofar as thinking
without the positing of categories and of concepts in general would be as
impossible as is breathing in a vacuum" (Schilpp, 674).
As I have tried to show in a number of case studies of scientific work,
stretching from the time of Kepler to that of Bohr and Einstein and to the
frontier work of today, we can recognize the existence of, and at certain
stages of scientific thinking, the necessity of postulating and using, precisely
such unverifiable, unfalsifiable, and yet not arbitrary conceptions, a class
to which I have referred as themata.
Among the themata which guided Einstein in theory construction are
clearly these: primacy of formal (rather than materialistic) explanation;
unity (or unification) and cosmological scale (egalitarian applicability of
laws throughout total realm of experience); logical parsimony and necessi-
ty; symmetry; simplicity; causality; completeness; continuum; and of course
constancy and invariance. It is themata such as these which explain in speci-
fic cases why he would obstinately continue his work in a given direction
even when testing against experience was difficult or unavailable. It explains
equally why Einstein refused to accept theories that were well supported
by correlation with phenomena but which were based on thematic presuppo-
sitions opposite to his own (as in the case of the quantum mechanics of
Niels Bohr's school).
This conception can be built into a modification of Fig. 1, to exhibit
the function of themata in the EjASE process. Fig. 7 shows a number of
possible leaps arising from E toward A, but only one (or a few) survive the
filtering action of the themata which a particular innovator h;ts adopted or in-
corporated into his conceptual processes. For example, the two conjectures
which Einstein raised to the status of postulates at the beginning of his 1905
relativity paper are thematically shaped presuppositions; they conform
Einstein's Model for Constructing a Scientific Theory 133

81.... .---r--,

8;
82 -+ ~---+,..---i

E
Figure 7

to the prior imposition of the requirements of large scale and egalitarian

applicability of laws, invariance, logical parsimony, and primacy of formal
explanations.

16 Confrontation of rival theories

The EJASE schema as now completed lends itself very well to repre-
sent the situation that arises when two different theories claim to handle
about the same experiential material. Thus Einstein often stressed that
his relativistic mechanics, at least in the early years, overlapped with Newto-
nian mechanics with respect to the range of testable experience, even though
the theories corresponded to "two essentially different principles" (1.0.,
p.273).
In analyzing controversies between rival explanations drawn from the
same ground of available experience - most recently the controversy bet-
ween Millikan and Ehrenhaft on the existence of the unitary charge of
electricity 12 - it has been evident to me that the difference of choices of the-
mata by the rivals explains a good deal of the details of their theories and of
the course of the controversy. In very abbreviated form, Fig. 8 will help to
make this point. On the left appears Al as the axiom system reached by
the first of two innovators. His themata are symbolized by eI' The sy-
stem of propositions gives rise to deductions S I, S~, S~, S~', as shown.
To most of these, there correspond observations (E I , E 2 , E 3 ) which can
be correlated with the deductions. As is usual, some deductions (st) re-
main, at least for the time being, without such "verifications," although
work may be in progress on just such a problem at that point.
The second innovator is represented by the system on the right side
of the figure. His axiom system A2 was reached through preliminary no-
tions passing through the constraint indicated by e2, his own set of the-
mata. The system A 2 as a whole is not so utterly different from A I on
the left side that there is not some overlap between the deductions made

12 G. Holton, The Scienllfic Imagination, chapter 2.

134 Gerald Holton

Figure 8

by the two innovators. Thus the deductions 52' 5; of the second person
refer to the same phenomena Eland E 2 , at least as far as can be deter-
mined at the time, as do 51 and 5~ . In addition, however, A2 allows deduc-
tion 5; for which there is no equivalent in the first system and which claims
to be correlated with ("borne out by") E4 .
Something like this is how a scientific dispute can continue for some
time. Al or A2 or both may, in the course of the debate, be progressively
modified, with corresponding changes in 51 and 52' Eventually, one or
the other of the two systems wins out, and this happens usually in one of
two ways. The two theory systems, separately, may come to a point of
development where there is no essential difference in the number and
types of phenomena (experimental evidences) which they can handle.
That is, perhaps by some ad hoc adjustment of A 2 , it too "can account
for" E 3 . If this situation persists for some time, a choice is made between
the two systems on the basis of the "appeal" of the fundamental presuppo-
sitions. This comes down to having the preponderance of the scientific
community making a choice on the basis of preferring the system of the-
mata 8 1 or 8 2 , Thus, in the early period when Einstein's relativity theory
could not be clearly distinguished from Lorentz's or Abraham's by any
significant differences in their testable predictions, Max Planck was driven
to exclaim in a scientific meeting, when pressed to confess why he believed
in Einstein's postulate system rather than its rivals: "I find it more conge-
nial" ["Mir ist das ... eigentlich sympatischer"J. 13
An alternative scenario is for one of the two systems to produce more
verifiable predictions of observable events than the other, and fewer (or no)
uncomfortable disconfirmations. Almost never is the situation so clean that
the inability of one theory system to handle a specific experiment produces
right away a decision in its disfavor. What is much more likely is that during
the period when attempts are made to account for apparent difficulties, the

13 M. Planck, "Prinzip der Relativitat," discussion at German Physical Society, 23 March

1906.
Einstein's Model for Constructing a Scientific Theory 135

balance of opinion swings toward one of the systems in favor of the other,
and the latter slowly fades from view without necessarily ever being "dis-
proved."

***
The model implied in Einstein's schema can be shown to be useful
in further, fruitful extensions, helping us to understand additional details of
the working of the scientific imagination. That is a task for another occasion.
By having gathered together Einstein's own expressed views, and having tried
to correlate them with the schema he himself proposed, we have found a
consistent prescription for one way the human mind may puzzle out the
order behind the appearances, and for communicating that perception to
others in a convincing manner.
Despite this grand ambition we must not think that Einstein in this
field, any more than in his science, attempted to impose an absolutistic
point of view. He was all too aware of the tentative state of understanding
in the field of scientific methodology. The spirit in which he proposed his
ideas is well conveyed in a passage in his Autobiographical Notes, imme-
diately after he has begun to give his answer to the question "What, preci-
sely, is 'thinking'?":
"With what right .~. the reader will ask - does this man operate so
carelessly and primitively with ideas in such a problematic realm without
making even the least effort to prove anything? My defense: all our thinking
is of this nature of a free play with concepts; the justification for this play
lies in the measure I Ubersicht lover the experience of the senses which we
are able to achieve with its aid" (Schilpp, p. 7).
As limited human beings confronting the seemingly endless, interlock-
ing puzzles of the universe, we can nevertheless hope to play - as in New-
ton's metaphor, with pebbles at the shore of a vast ocean. If we do it well,
that play can yield the most highly desirable kind of knowledge: a survey
(overview, Ubersicht) of the world of nature that grants us the perception of
order guiding the phenomena in their infinite, individual variety, and in their
inexhaustible interactions with one another.

Acknowledgement

I wish to express my indebtedness to Miss Helen Dukas and the Estate of Albert
Einstein for help and for permissions to cite from the publications of Einstein. I am
happy to acknowledge research support by grants from the National Science Founda-
tion and the National Endowment for Humanities.
136 Gerald Holton

Principal References

Albert Einstein, Ideas and Opinions, new translations and revisions by Sonja Barg-
mann (New York: Crown Publishers, Inc., 1954, and, with somewhat different pagina-
tion, New York: Dell Publishing Co., 1954). These essays are based on Mein Weltbild,
a collection of Einstein's essays, edited by Carl Seelig, and other sources.
Among the essays specifically cited in this article are the following (all in Crown edi-
tion): "Remarks on Bertrand Russell's Theory of Knowledge," 1944 (pp. 18-24)
"A Mathematician's Mind," 1945 (Letter to J. S. Hadamard) (pp. 25-26)
"Principles of Theoretical Physics," 1914 (pp. 220-223)
"Principles of Research," 1918 (better: "Motive of Research") (pp. 224-227)
"What is the Theory of Relativity?," 1919 (pp. 227-232)
"Geometry and Experience," 1921 (pp. 232-246)
"On the Method of Theoretical Physics," 1933 (Herbert Spencer Lecture) (pp. 270-276)
"The Problem of Space, Ether, and the Field in Physics," 1930-1934 (pp. 276-285)
"Physics and Reality," 1936 (pp. 290-323)
"The Fundaments of Theoretical Physics," 1940 (pp. 323-33 5)
"On the Generalized Theory of Gravitation," 1950 (pp. 341-356)
Paul Schilpp, editor, Albert Einstein: Philosopher-Scientist (Evanston, Illinois:
The Library of Living Philosophers, 1949). This book includes (pp. 3-94) Einstein's
"Autobiographical Notes," a number of excellent essays on Einstein's work by scientists
and philosophers, and (pp. 665-688) a set of supplementary comments by Einstein. The
"Autobiographical Notes" are scheduled to be published in 1979 as a separate book,
under the editorship of P. A. Schilpp and Otto Nathan.

Other Reading and References

Albert Einstein and Leopold Infeld, The Evolution of Physics (New York: Simon
and Schuster, 1938).
Albert Einstein, Lettres aMaurice Solovine (Paris: Gauthier-Villars, 1956).
Philipp Frank, Einstein, His Life and Times (New York: A. A. Knopf, 1947).
Banesh Hoffmann, Albert Einstein, Creator and Rebel, with the collaboration
of Helen Dukas (New York: The Viking Press, 1972).
Gerald Holton, The Scientific Imagination: Case Studies (New York and Cam-
bridge: Cambridge University Press, 1978), chapters 1-3.
Gerald Holton, Thematic Origins of Scientific Thought: Kepler to Einstein (Cam-
bridge: Harvard University Press, 1973), chapters 5-10.
 137

Einstein's Treatment of Theoretical Concepts

Bernulf Kanitscheider

I.

In all realms of nature science has repeatedly been confronted with

the problem of introducing theoretical concepts which do not deal with
observable entities for a coherent description of phenomena. In view of
this situation there have always been two methodological procedures
which are clearly reflected in the principles of theory construction. One
method starts at the level of the phenomena and extrapolates into those
fields that are not accessible empirically. This method tries to do as much
as possible without theoretical constructs, i. e. without predicates and
relations which have no direct correlates in experience. If theoretical
constructs cannot be avoided, one tries to introduce them as auxiliary
concepts without semantic reference, which serve but as syntactic inter-
mediaries between terms of empirical meaning. Opposed to this is the
other strategy which uses without hesitation terms that refer to non-per-
ceptible entities. According to this method the linguistic reference is justified
if results of logical deductions have been validated in the field of experience.
In this case it is stated that the reference of the theoretical term to the
hidden object has been validated.
From a historical point of view toc second method is certainly older.
The earth-quake hypothesis of Thales 1 and Anaximander's theory on the
suspension of the cylindrical earth in the universe are obviously explanatory
approaches where items of the visible world shall get understandable by
means of hypothetical but supposedly real acting causes of the phenomena.
The first case uses as hypothetical structure the floating earth on the okeanos,
the latter the intrinsic structural symmetry of the universe.
The origin of the first method is probably the failure of certain theories
which adopt an inner mechanism as an explanation of the phenomena. The
failure of the theory of the homocentric spheres of Eudoxos, Kalippos, and
Aristotle - particularly because it could not account for the apparent
brightness of the planets - raised doubts in the methodological correctness

1 W. Capelle: Die Vorsokratiker, Hamburg, 1953, p. 70, no. 3.

2 W. Capelle: l. C., p. 79, no. 13.
138 Bernulf Kanitscheider

of the search for invisible, but, according to the Platonic axiom, real mecha-
nisms which save the phenomena. From a logical point of view it is always
justified, when a theory fails, to doubt its methodological basis. The doubt
concerning the Aristotelian principle to use in astronomy representational
theories that can be interpreted realistically, led to Ptolemai's phenomenalistic
approach, i. e., to a theory which used the empirical material for prognosis
in an instrumentalistic way, but could not produce true explanations of the
actual movements of the celestial bodies 3 • This methodology was changed
by Copernicus. He took up Aristarch's and Seleukos' bold conjecture that
the queer paths of the planets are accidental features due to the terrestrial
human perspective. They can be understood causally, if one can make up
one's mind to accept unobservable processes, namely, the three forms of
the earth's motion.
This controversy is even more clearly reflected in the two methodologi-
cal positions in the history of atomism. Democrites tried to formulate his
epistemological idea that the visible world can be explained through ac-
cretion of a small number of transempirical objects. The spatial arrange-
ment and the kind of connection between the micro-objects were to re-
produce the known structures on the level of perception. The development
of the atom-hypothesis shows clearly that the mechanism of the connection
of the elementary constituents, as well as the linkage of the atomic complexes
with the level of phenomena were so poorly defined that its methodology
could not be trusted. The evidence became more convincing, however, when
Berzelius discovered the electrical nature of the chemical binding forces and
Dalton confirmed the connection with the (chemical) phenomenological level
through the law of multiple proportions. But even then, only indirect hints
at the invisible microscopic level resulted therefrom; they can only be referred
to as derived but not as direct evidence. 4 Therefore the opponents of the
atom-hypothesis - grounding their arguments on a positivistic epistemology -
could easily make a case for treating the atoms as useful fictitious concepts
having only the syntactical auxiliary function (as mentioned earlier) to link
certain sense data but do not have any (referential)meaning.
How strong the influence of the empiricist methodology was at that
time can be inferred from the fact that even Planck doubted initially the
reliability of the atomic hypothesis. He changed his mind only when he had
to make use of the probabilistic assumption of the molecular kinetic theory
in the derivation of his radiation law. 5 Only a few scientists were able to
evade the influence of the Mach-Kirchhoff descriptivism which emphasized

3 Norwood Russell Hanson: Constellations and Conjectures, Dordrecht, 1973, p. 89.

4 Only after the invention of the field-ion-microscope one could speak about direct
evidence in connection with the atom hypothesis.
5 M. Jammer: The conceptual development of quantum mechanics, New York, 1966,
p.19.
Einstein's Treatment of Theoretical Concepts 139

economy of thought, ontological parsimony and mistrust for unobservable

quantities. Among those who were able to resist the then current trend
towards positivism was Ludwig Boltzmann. He was completely aware of the
hypothetical character of theoretical entities:
"The existence of these minute single bodies (the molecules) which
interact to constitute the bodies which we conceive with our senses, is only a
hypothesis, just as it is a hypothesis that what we see in the sky is caused
by cosmic bodies immensely large and distant, and as it is, after all, a hypoth-
esis that other human beings capable of feeling pleasure and pain exist
outside myself. "6
The reason for Boltzmann's attitude which was opposed to contemporary
trends can be explained: His research program was the creation of a unified
mechanical world view. His epistemology was embedded in a comprehensive
idea for which it was unbearable that two classes of phenomena exist, i. e.
reversible phenomena described by mechanics and electrodynamics, and
irreversible processes described by thermodynamics. For the sake of theo-
retical unification Boltzmann had to use the molecular kinetic hypothesis.
He used this hypothesis not in a purely fictitious sense, but used the concept
of the atom as postulated entity, the existence of which is justified by the
validation of the theory (as shown by the quotation given above). In doing
so he had to use a presupposition which was analyzed and justified philo-
sophically only much later from a more refined semantical point of view.
This presupposition may be called the fundamental postulate of hypothetical
realism: the number and the type of objects to which a theory refers can be
different from the number and type of phenomena which serve to validate
the theory. This difference will turn up nearly always in the case of high
ranking theories with great prognostic and explanatory power. A physical
theory can therefore contain statements about classes of empirically in-
accessible objects, as long as the semantics of the theory contains statements
linking the hidden entities and the processes in which they are involved with
phenomenological tests. This goes beyond the positivist point of view in the
sense that uncontrollable metaphysics does not begin with statements about
invisible quantities. It begins only when one does not succeed in connecting
the invisible by means of a translucid interaction-mechanism, which can
be tested in an other theoretical context, with the level of observations.

II.

Of special relevance for the philosophical examination of Einstein's

physical ideas is the fact that Einstein was influenced in his early years
not only by Mach but also by Boltzmann's kinetic theory of gases, from

6 L. Boltzmann: Populare Schriften, Leipzig, 1905, p. 30.

140 Bernulf Kanitscheider

which he gained his knowledge on statistical mechanics. Especially in Ein-

stein's early papers on statistical physics one finds a concentration on the
same type of scientific objects as in Boltzmann's research program, and
furthermore also the same methodology, as Einstein confirmed later himself:
"My major aim in this was to find facts which would guarantee as much as
possible the existence of atoms of definite finite size."7 In order to verify
this assertion it is best to start from Einstein's early papers on statistical
physics. The problem raised in his paper on Brownian motion 8 was the
search for links between phenomenological thermodynamics and the level of
invisible microphenomena introduced by Boltzmann in order to increase the
explanatory power and to unify the description given by the theory. According
to thermodynamics there is the static final equilibrium state of any thermo-
dynamic system. This state is considered, from the point of view of the
molecular kinetic hypothesis, as a stationary state of the system in which the
microprocesses take place at such a rapid rate that they usually cannot be
observed on the level of phenomena. The question arises whether a link can
be found at least in some special cases. Einstein found a link, surprisingly
without knowing that Robert Brown had investigated already in 1827 the
"Zitterbewegung" of small particles suspended in a fluid. Brown had also
noticed that the fluctuations increase when the temperature is raised and
decrease when larger particles are suspended. Thus Einstein did not endeavor
to explain a known fact with the help of a new molecular kinetic theory of
matter, but he tried to find an indicator for a new level of reality postulated
hypothetically. For this purpose he studied the unobservable displacement x
of a small particle (it has to be much larger than the molecules, however)
and derived from this, by means of a logical construction, the rms-displace-
ment t = y'X2 .9 The decisive step of linking the two different levels is then
an identification of a quantity formed mathematically out of t with the co-
efficient of diffusion D which can be observed macroscopically: D = T,en
where T is the time interval considered. The decisive progress in this research
program formulated for the first time by Democrites, is the creation of a
quantitative link between both levels: the quantity D which occurs in the
diffusion equation (i. e. in a macroscopic theoretical relation) is connected in
a quantitative way with the size a of the atoms by D = kTI6 7r 77 a (77 is the
viscosity and k is Boltzmann's constant). This relation enabled Jean Perrin to
determine the Avogadro number with unprecedented accuracy. Thereby an

7 A. Einstein: Autobiographical Notes, in; P. A. Schilpp (ed.); Albert Einstein: Philo-

sopher-Scientist, The Library of Living Philosophers, 2nd edition, New York, 1951,
p.47.
8 A. Einstein: Ober die von der molekularkinetischen Theorie der Wiirme geforderten
Bewegungen von in ruhenden Fliissigkeiten suspendierten TeiIchen, Ann. Phys. 17
(1905), p. 549-560.
9 Cornelius Lanczos: The Einstein Decade, London, 1974, p. 59.
Einstein's Treatment of Theoretical Concepts 141

effective check of the theoretical entities introduced hypothetically was

possible. The differences in opinion in the opposition against the realistic use
of theoretical entities can be seen from the fact that Ostwald, one of the
opponents of atomism, was convinced by Einstein's paper while Mach's
sceptical attitude was shaken only when he saw the light flashes of a-particles
on a scintillation screen, 10
With respect to the acceptance of theoretical entities, one must distin-
guish between a psychological and a logical level. From a logical point of
view there can be no gradual transition from an instrumentalist to a realist
interpretation of theoretical entities. Only a simple alternative is given here,
since no quantity can playa role between real and fictitious entities: no
different degrees of "reality" exist. Psychologically, such a gradual transition
takes place when the couplings between the theoretical and the empirical
level become more numerous and increase the confidence that the theoretical
concepts are not purely fictitious. Furthermore, one has to be aware that
nobody can be forced to agree if he does not accept the light flashes seen
in radioactive decays to be the result of ultimately invisible constituents of
matter, but considers these constituents to be linguistic auxiliaries only,
introduced for the sake of ordering the flashes of light. The causal relations
are restricted in this case to relations between these phenomena, but are
not applied to the origin of the phenomena. The confidence in the realistic
interpretation of theoretical entities is enhanced in a psychological sense by
the concordance of empirical results obtained with different theoretical
methods and different experimental setups. The agreement between mole-
cular diameters determined on the one hand by the law of radiation,I1 and
on the other hand from Brownian motion is a very surprising and irreducible
fact for an advocate of the fictionalistic position. From a realistic point of
view this agreement is a natural consequence of the basic assumption that
the same underlying entities are present in both cases. The agreement ob-
tained with many logically independent methods of measuring the Avogadro
number and molecular diameters must make a fictionalist believe in a uni-
versal conspiracy of the phenomena if he wants to deny their independent
common origin. Nevertheless, one has to be aware that seen from a logical
point of view nobody can be forced to explain the surprising agreement
obtained. Only epistemological arguments tending in the direction of a
coherent model of scientific knowledge can be decisive here. 12

10 S. G. Brush: A History of Random Processes I. Arch. Hist. Exact Sci. 5 (1969), p. 35.
idem: Mach and Atomism Synthese 18, 213 (1968), p. 208.
11 M. Planck: Zur Theorie des Gesetzes der Energieverteilung im Normalspektrum, in:
Dokumente der Naturwissenschaft, ed. by A. Hermann, vol. 12, Stuttgart, 1969,
p.25-33.
12 Cf. B. Kanitscheider: Die Philosophie der modernen Physik, Section II: Sprache und
Erkenntnis, Darmstadt, 1979.
142 Bernulf Kanitscheider

The role of hypothetical postulates in Einstein's thinking is especially

apparent in his papers on the radiation problem. 13 There is hardly any
other paper in physics in which the constructive elements are more remote
from the phenomena. Einstein was well aware of the audacity of his
assumptions, as can be seen from the cautious wording "heuristic point
of view" instead of "hypothesis". It is one of the greatest ironies in the
history of science that Planck considered the quantum hypothesis to be a
speculation shooting beyond its aim, as he wrote in his letter of recommenda-
tion on the occasion of Einstein's enrollment in the "Preussische Akademie
der Wissenschaften"14. It has been argued that the line of thought which led
to the light quantum-hypothesis is not in agreement with the hypothetical-
deductive scheme usually considered in the theory of science, and has to
be subsumed under a different strategy called "deduction from the phenom-
ena".15 A closer analysis shows, however, that the relation between the
theoretical results and the phenomena is of the very indirect form char-
acteristic for hypothetical-deductive reasoning. It is certain that the heuristic
basis of Einstein's work was not the desire to solve the riddle of the photo
effect, which was a very erratic and irritating element in the field theory
of electro-magnetism especially since the experiments of Hertz and Hallwachs
(1888) and the quantitative analysis of Lenard. Einstein's starting point
was a rather general symmetry consideration of a more natural philosophical
kind: there seems to be a fundamental difference between the conceptual
structure used for the description of solids and gases and the concepts
needed for the explanation of electromagnetic phenomena. Only a finite
number of quantities is needed to fix the state of the discrete atomic con-
stituents of matter, while a set of continuous functions is necessary for this
purpose in electro-magnetic theory. It was the dualism between particles
and fields which seemed unacceptable to Einstein. The question arose
whether Maxwell's theory described only the average bei!avior of quantities,
while the elementary acts of emission and absorption of light are discontin-
uous and closer to the particle concept than the field concept.
A short remark may be added here. In his autobiographical notes Ein-
stein confirmed the supposition that one of his guiding ideas had been
ontological monism. 16 Lorentz had changed electrodynamics in a manner
which Einstein considered to be a fundamental improvement. According to
Lorentz it is space and not matter which is the carrier of the field. Atomistic

13 Ober einen die Erzeugung und Verwandlung des Lichtes betreffenden heuristischen
Gesichtspunkt, Ann. Phys. 17 (1905), p. 132-148.
14 Cf. Dokumente der Naturwissenschaft, ed. by A. Hermann, vol. 7, Stuttgart, 1965,
p.13.
15 Jon Dorling: Einstein's Introduction of Photons. Argument by Analogy or Deduction
from the Phenomena, Brit. J. Phil. Sci. 22 (1971), p. 1-8.
16 A. Einstein: Autobiographical Notes, I. c.
Einstein's Treatment of Theoretical Concepts 143

matter is the site of the charges only, and the field is between the charged
matter. Einstein considered it unsatisfactory, however, that both the concept
of material points and of continuous fields were used as primitive terms
even within electromagnetism:
"If one views this phase of the development of theory critically, one is
struck by the dualism which lies in the fact that the material point in
Newton's sense and the field as continuum are used as elementary concepts
side by side. Kinetic energy and field-energy appear as essentially different
things. This appears all the more unsatisfactory inasmuch as, according to
Maxwell's theory, the magnetic field of a moving electric charge represents
inertia. Why not then total inertia? Then only field-energy would be left,
and the particle would be merely an area of special density of field-energy.
In that case one could hope to deduce the concept of the mass-point
together with the equations of the motion of the particles from the field
equations - the disturbing dualism would have been removed."17
This ontological dualism cannot be avoided within the framework of
a linear theory based on the principle of superposition. Thus his metaphysical
search for unification is likely to be one of the sources of Einstein's pref-
erence for non-linear theories in which not only the properties of the basic
elements but also their interaction can be described.
In order to arrive at the light quantum-hypothesis Einstein 18 used
three theoretical laws: the statistical law WI = (vlvo)n [which gives the
probability to find n particles contained in a volume Vo in the smaller
volume v], Wien's law p (v) = <xv 3 e- hvlkT [the limit of Planck's law for large
values of vlT] and the Boltzmann relation between entropy and probability
S = k ·log W [here statistical assumptions enter, e. g. in the collision term]. A
number of logical steps lead from these theoretical elements to an expression
for the probability W 2 = (vlvo)Elhv of finding the total energy of radiation
at any moment in a partial volume v. From the structural analogy of the
expressions WI and W 2 he concludes the (semantic referential) identity of
the exponents: E = n h v. I n the key sentence of his paper Einstein still uses
the fictionalistic interpretation of the light quanta: "Monochromatic radiation
of low density ... acts from the point of view of thermodynamics, as if it
consisted of mutually independent energy quanta of the size h v." 19 He then
explains some effects which were previously not understood (the dependence
of the energy of the electrons emitted in the photo-electric effect on the wave-
length, Stokes' rule for fluorescence, the release of secondary cathode rays

17 A. Einstein: Autobiographical Notes, I. c., p. 37.

18 For details cf. Martin}. Klein: Einstein's First Paper on Quanta, The Natural Philo-
sopher 2 (1963), p. 57~86.
19 A. Einstein: Ober einen die Erzeugung ... , in: Dokumente der Naturwissenschaft,
vol. 7, p. 37.
144 Bernulf Kanitscheider

with unchanged velocity), and thereby strengthens his belief in the auto-
nomous existence of light quanta.
It becomes even more apparent that photons were new theoretical
entities sui generis which were to be given their own position within physical
ontology when Einstein studied the question of statistical fluctuations in a
radiation field. 20 It turned out that the restrictive interpretation of the
photon concept advocated by Planck 21 was untenable. Planck intended to
use the photon concept only for processes of emission and absorption and
considered it to be a stop-gap which had to be used for epistemic reasons
such as the far-reaching ignorance of the interaction between matter and
the radiation field. It turned out to be necessary, however, to accept the
permanent existence of photons and the universal occurrence of radiation
in multiples of h v. Otherwise one cannot ullderstand the fluctuation equation
for the energy (e 2 )112 dv = (h v + c 3 pl8 7r v2 ) 7rV d v, which Einstein derived for
a mirror which can rotate freely within a radiation field cavity. This result is
remarkable in two respects: it shows that fluctuations consist, on the one
hand, of one component following from the classical wave theory of light;
i. e. fluctuations of the radiation pressure exist because wave packets inter-
fere with one another. On the other hand, there is an extra term which is
due to the reflection of light quanta, and only the combination of both
effects leads to the correct average energy! kT of the mirror. "This way of
looking at the problem showed in a drastic and direct way that a type of
immediate reality has to be ascribed to Planck's quanta, that radiation must,
therefore, possess a kind of molecular structure in energy, which of course
contradicts Maxwell's theory. "22 The fluctuation equation is remarkable in
still another respect. It contains the first hint to the phenomenon of duality
which was later the starting point for many interpretations of quantum
theory, since neither the elements of the particle theory nor of the wave
theory can be eliminated from the fluctuation expression.
It is often stated that Einstein's early work was influenced by the
philosophy of science prevailing at that time towards phenomenalism,
operationalism, or positivism, while his later work, starting with general
relativity, was dominated by critical realism, a sense of objectivity, and
the necessity of speculation. The examples given above from Einstein's
early work lead to the impression that no such break in his methodology
exists. The alleged change in Einstein's epistemological position was often
noted and considered to be negative, possibly seen in connection with the

20 A. Einstein: Zum gegenwartigen Stand des Strahlungsproblems, Phys. Zs. 10 (1909),

p. 185-193.
21 M. Planck: Die Gesetze der Wiirmestrahlung und die Hypothese der e1ementaren
Wirkungsquanten, in: idem: Die Theorie der Strahlung und der Quanten, Halle, 1913,
p.77-94.
22 A. Einstein: Autobiographical Notes, I. c., p. 51.
Einstein's Treatment of Theoretical Concepts 145

failure of the field theoretic program which Einstein worked on in his later
years. In this negative way the change of position was interpreted mainly by
Bridgman:
''That in his conviction of the possibility of getting away from any
special co-ordinate system, in his conviction of the fruitfulness of so doing,
and in his treatment of the event as something primitive and unanalyzed, he
has carried into general relativity theory precisely that uncritical, pre-
Einsteinian point of view which he has so convincingly shown us, in his
special theory, conceals the possibility of disaster. "23
In the same way it is still seen by many theoreticians:
"The founder of the 'dualistic' interpretation was Albert Einstein,
more accurately the younger Einstein, who did not speculate as he did in
his later years, but analyzed the empirical data with great acuteness and
derived unassailable conclusions therefrom. "24 Is the assertion of a method-
ological break really correct?
It is true that an operationalistic language is used in the first five
paragraphs of the paper "On the Electrodynamics of Moving Bodies", i. e.,
in the so-called kinematical part of his fundamental paper on special rel-
ativity. In the short introduction Einstein makes some statements, however,
on his heuristic principles from which a completely different epistemological
position becomes apparent. He is irritated mainly by the asymmetries of
Maxwell's theory, which seem to have no counterpart in experience, and
are present when the theory is applied to moving systems. The theory deals
with the interaction between a magnet and a conductor in a different way,
depending whether the conductor or the magnet is moving with respect to
absolute space (represented by the ether), although the relative state of
motion is the same. It is true that Einstein also mentions an experiment
- the unsuccessful attempts to measure the motion of the earth relative to
the ether 25 - in support of the conjecture that an absolute system of ref-
erence is singled out neither by mechanics nor by electrodynamics. The
decisive step is taken, however, when Einstein raises the conjecture that the
fundamental laws of mechanics and electrodynamics have the same form in
all systems of reference which are moving uniformly with respect to one
another, to the state of a principle. From this principle and the postulate

23 P. W. Bridgman: Einstein and Operationalism, in: P. A. Schilpp (ed.): Albert Ein-

stein, I. c., p. 354.
24 M. BornlW. Biem: Dualismus in der Quantentheorie, Phil. Nat. 10,4 (1968), p. 411.
25 A. Einstein: Zur Elektrodynamik bewegter Korper, in: idem: Das Relativitatsprinzip,
Darmstadt, 1974, p. 26. - Exact historical analyses have proved that it was not the
Michelson experiment which played the decisive heuristic role but that Einstein
thought at the phenomenon of the aberration and Fresnel's ether-drag hypothesis.
Cf. G. Holton: Einstein, Michelson and the 'Crucial Experiment', Isis 60 (1969),
p.I33-197.
146 Bernulf Kanitscheider

of the constancy of the speed of light he deduces a new kind of electro-

dynamics which is free of internal contradictions. Retrospectively Einstein
considered his starting point to be the solution of a logical dilemma. 26 Three
assumptions are mutually exclusive:

(1) The co-ordinates of two inertial systems are related by Galilei trans-
formations.
(2) The laws of nature are independent of the choice of the inertial system.
(3) The velocity of light has the same constant value in every inertial
system.

It is well known that his solution to this dilemma consisted in a change

in assumption (1). He replaced the Galilei transformation by the Lorentz
transformation. Since electrodynamics was already Lorentz-invariant, only
a reformulation of mechanics was needed. The laws of both fundamental
physical disciplines could be inferred mainly from symmetry considerations
in connection with a unified group of transformation. These symmetry
considerations had consequences which could be checked on the level of
phenomena:
"With the given physical interpretation of co-ordinates and time, this is
by no means merely a conventional step, but implies certain hypotheses con-
cerning the actual behavior of moving measuring-rods and clocks, which
can be experimentally validated or disproved. "27
Einstein's method for constructing theories started thus by postulating
principles which were not intuitively obvious ([ 3] especially contradicts
the immediate intuition). From these principles deductive consequences are
inferred and dressed in an operational language in order to be more readily
understandable, that is, for didactic reasons. In order to support this con-
tention one can give psychological reasons: the new and uncustomary
relations between old concepts like space and time had to be presented in
a form which was as evident and transparent as possible. That the kinematical
part of Einstein's great paper on special relativity is presented in a didactic
way can also be supported by the fact that the entire contents of special
relativity can be transformed in a non-operational language without any
loss of meaning. 28 It would be a grave hindrance for the applicability of the
theory if such a transformation were impossible and the operational language
were to be a part of the physical content which could not be eliminated.
In that case the theory could only be applied to a world in which observers
with rigid measuring rods and isochronous clocks were present in all of
space - a world which differs essentially from the one in which we live.

26 A. Einstein: Autobiographical Notes, 1. c., p. 57.

27 A. Einstein: Autobiographical Notes, 1. c., p. 57.
28 Cf. W. Rindler: Special Relativity, Edinburgh, 1960.
Einstein's Treatment of Theoretical Concepts 147

Actually it is easily possible to talk about the simultaneity of events here

and, for example, on Sirius without introducing an observer about whose
existence we know nothing, who sits on Sirius and emits light flashes. 29 The
assertion given above can also be supported by biographical reasons, since
Einstein himself rejected the normative generalization inferred by Bridgman
from the didactical formulation of special relativity:
"In order to be able to consider a logical system as physical theory
it is not necessary to demand that all of its assertions can be independently
interpreted and 'tested' 'operationally'; de facto this has never yet been
achieved by any theory and cannot at all be achieved. In order to be able
to consider a theory as a physical theory it is only necessary that it implies
empirically testable assertions in general. "30
Furthermore, other reasons have since been found why special relativity
should not be considered to be a theory for which the semantic reference
consists of measuring rods and clocks.31 Firstly, the theory applies not only
to the macroscopic world in which measuring instruments exist, but also to
the world of elementary particles where measuring rods and clocks cannot
exist for nomological reasons. If the operational prescriptions were to be
taken seriously, the theory had to contain linguistic elements describing the
mechanism and the function of the instruments. This is neither the case in
special relativity nor in quantum mechanics, which contains no syntactic
blank to the omnipotent observer and his instruments. In spite of this,
there is a true core to operationalism: even if the assumptions about the
functioning of instruments are not needed in order to express the statements
about the world contained in a theory, they are necessary in order to follow
the causal chain during test instances which determine the truth value of
statements concerning the section of reality described by the theory. T6rne-
bohm has formulated the logical independence of the semantic interpretation
from the test program of a theory very precisely by considering the onto-
logical reference to be a mapping. 32
A mapping over a region X which is part of the world is an X-ological
mapping; a mapping over instruments which create information on X is an
X-ometric mapping. In the case of special relativity one has an atlas of X-

29 j. j. C. Smart: Between Science and Philosophy, N. Y., 1968, p. 140. Reichenbach

had arrived at the same conclusion, too: "In a logical exposition of the theory of
relativity the observer can be completely eliminated." (H. Reichenbach: The Philo-
sophical Significance of the Theory of Relativity, in: P. A. Schilpp (ed.): Albert
Einstein, I. c., p. 295.)
30 P. A. Schilpp (ed.): Albert Einstein. Reply to Criticisms, in: Albert Einstein, I. c.,
p.679.
31 H. Tornebohm: Aspects of the Special Theory of Relativity, in: E. Laszlo and
E. B. Sellon (eds.): Vistas in Physical Reality, New York, 1976, p. 31-62.
32 H. Tornebohm: I. c., p. 32.
148 Bernulf Kanitscheider

ological maps in which X denotes space-time properties and series of events.

Thereby the realm of special relativity is fixed in a non-operational way.
Events in space-time and their causal relations exist also when there happens
to be no observer present who synchronizes clocks with light rays and
transports rigid rods. The use of these instruments lets the system of X-
ological maps become manifest as test instances in certain discrete places.
Even the relativity of simultaneity, Einstein's result which is most revolu-
tionary in its philosophical implication, can be formulated in an objectivistic
language. The simultaneity of events a, b, c is an equivalence relation G (x, y)
such that G (a, a), G (a, b) -+ G (b, a), and G (a, b)" G (b, c) -+ G (a, c). The
relativity of simultaneity invalidates the transitivity of G (x, y) for systems
of reference in relative motion, it restricts the transitivity to the case in
which the system of reference for which G (a, b) is valid, coincides with the
one for which G (b, c) is valid. 33 In this manner the relativity of simultaneity
has been formulated in an objective way, which is, however, not testable.
If one wants to know whether the statement given above is true, a geo-
chronometric re-interpretation is necessary in which concrete types of
atomic clocks in moving systems of reference assume the role of test instances
as, for example, in the experiment of Hafele and Keating. 34

III.

In the preceding chapters we have followed Einstein's advice to infer

the epistemological principles of a physicist from his actual scientific work
and not from his own philosophical reflections about it. It is interesting,
however, to compare the epistemological structures contained in Einstein's
work and his own reconstruction of these structures with the philosophical
advice offered in those years concerning the genuine scientific method. It
is remarkable that the problem of the status of the theoretical entities was
not a central problem in the early days of logical empirism. One was pre-
occupied with the separation of philosophy from metaphysics, which was
considered to be obscure. The rigorous trait of the philosophy of science
during the pre-war years can be explained in this way and one can also
understand why the members of the Vienna Circle initially considered
highly developed natural sciences to present the ideal model for rigor in
theory construction, despite the fact that the scientists themselves were
much more liberal. Their views and their practical approach towards the
handling of meaning criteria for theoretical concepts and statements as well

33 K. R. Popper: Intellectual Autobiography. in: P. A. Schilpp (ed.): The Philosophy

of Karl Popper, La Salle (Ill.), 1974, p. 77.
34 Cf. R. SexliH. K. Schmidt: Raum-Zeit-Relativitat, Hamburg, 1978, p. 39.
Einstein's Treatment of Theoretical Concepts 149

as their admission of non-empirical boundary conditions for the selection

of theories was not as rigorous as the philosophy of science wanted to have
it. It appears to be enlightening to contrast here the intellectual development
of the most important representative of logical empiricism with Einstein's
methodological reflections. It is instructive to compare the program on
which Carnap's "Logischer Aufbau der Welt" ist based 35 with a statement
by Einstein (1933) 36 :
"The structure of the system is the work of reason; the emprical con-
tents and their mutual relations must find their representation in the con-
clusions of the theory. In the possibility of such a representation lies the sole
value and justification of the whole system, and especially of the concepts
and fundamental principles which underlie it. These latter, by the way, are
free inventions of the human intellect, which cannot be justified either by
the nature of that intellect or in any other fashion a priori. "
According to Einstein the conceptual reconstruction of the world
starts from a free choice of basic terms and their connections, which have
to be justified afterwards by their empirical confirmation. According to
Carnap an observational basis (rock bottom of knowledge) is presupposed,
which is considered secure and from which highly organized knowledge is
built up by logical construction. Mach's sensualism, Russell's logical atomism
and Wittgenstein's thesis that all complex statements are truth functions of
elementary statements, have contributed to this concept. 37 Only later did
a more liberal form of empiricism prevail in the Vienna Circle and it entailed
different understanding of the relation between the epistemologically
abstract key-terms of scientific theories and the concepts of the empirical
basis which refer to the observable properties of material bodies. 38 The
methodological ideal of reduction was maintained, but the admissible
relations between empirical and theoretical concepts became more indirect.
Explicit definitions were replaced by reduction sentences but even those
were soon insufficient to guarantee the definitory substitution of theoretical
terms. The importance of unobservable quantities was recognized only
later. 39 Carnap reconstructs scientific disciplines as formalisms, the axioms
of which are the fundamental equations of the field considered. The formalism
is not interpreted directly but as a network of theoretical concepts which are

35 E. Carnap: Oer logische Aufbau der Welt, Berlin, 1928.

36 A. Einstein: On the Methods of Theoretical Physics, Ill: A. Einstein: Essays in
Science, New York, 1934, p. 15.
37 R. Carnap: Intellectual Autobiography, in: P. A. Scbilpp (ed.): The Philosophy of
Rudolf Carnap, La Salle (111.),1963, p. 57.
38 R. Carnap. Testability and Meaning, Phil. Sci. 3,4 (1936), p. 419-471; 4, 1 (1937),
p.1-40.
39 R. Carnap: Foundations of Logic and Mathematics, International Encyclopedia of
Unified Science 1,3, Chicago, 1939.
150 Bernulf Kanitscheider

connected by axioms (freely floating system). Using the fundamental con-

cepts one can define new concepts, and only some of these can be connected
with the observables by means of semantic rules. This is undoubtedly a
first form of the two-language model in which a fully interpreted empirical
language helps to make a theoretical language consisting of a network of
abstract terms partially understandable by means of rules of correspondence.
The usefulness of the two-language model is rather limited since the
distinction between empirical and theoretical terms is mostly conventional.
There is, for instance, a continuous transition from the world of objects
seen through normal window glass, through glasses with two dioptries,
through a normal microscope, to the use of an electron or field-ion micro-
scope, in which even single atoms can be made visible. It seems to be highly
arbitrary, however, to assign a theoretical status to objects seen above a
certain resolving power of the instrument and an empirical status to the
rest. An optical theory is needed in every case even for the view through a
plain glass window or with the naked eye in order to explain why this gives
reliable information about material objects. Theoretical reasons lead to
further arguments against the unequivocal classification of the two groups of
concepts. 40 An example of this is the present theory of chemical bond. It
asserts a steady transition from small molecules like H2 to larger ones like
fatty acids, polypeptides, or virus, to extremely large molecules like diamonds
or plastics. The objects of the latter group are directly observable although
they are genuine single molecules. It would be a highly arbitrary act to draw
a border-line between a large protein molecule, which can be observed only
with an electron microscope and a polymer which is seen in an optical
microscope. This example shows that a classification of concepts into
empirical and theoretical ones is meaningless. It shows furthermore that the
assignment of a syntactic auxiliary status to invisible micro-constituents of
matter cannot be maintained.
With reference to Einstein's own work there is' an example which
shows clearly that the distinction mentioned above depends on historical
circumstances and is of little relevance. When Einstein constructed his theory
of gravitation there was certainly no object more epistemologically abstract
than the Riemannian curvature R~ro . Meanwhile, methods have been found to
measure the curvature properties of space-time and to connect them causally
with visible phenomena in the same way as an electromagnetic field is tested
with the help of test charges. In order to measure the space curvature in
some region, one observes the behavior of two neighboring test particles and
derives the curvature directly from the geodesic deviation
D2 ~ex ex dx f3 dx o
--+R _~r __ 0
Dr2 f3r fi dr dr -

40 G. Maxwell: The Ontological Status of Theoretical Entities, Minnesota Studies,

vol. III, 1962, p. 9.
Einstein's Treatment of Theoretical Concepts 151

This equation describes the relative acceleration of the two particles. A

special instrument, the gravitational gradiometer, has been constructed for
the measurement of static and slowly varying curvature properties. Fast
variations of the curvature can be determined with the help of gravitational
wave-antennas. 41 Certainly there are a number of theoretical steps to be taken
in connecting the torsion of orthogonal hands of the gradiometer with the
observational statement: "The Riemannian curvature produced by a mountain
with the height of 2 km in a distance of 15 km is roughly 10- 30 em -2."
These steps are, however. not fundamentally different from those leading
from a statement about an observation made within glasses with two-dioptries
and the corresponding macroscopic object. In this case the theoretical steps
are provided by geometrical optics.
In the middle of the fifties philosophers started to be convinced that
theoretical statements can never be translated into an empirical language.
On the contrary, the freedom in the formulation of concepts and the ex-
planatory and prognostic power connected with this would be severely
limited 42 , if any such type of empiricist linguistic reduction were demanded.
It is thus not surprising to find the following statement in the later writings
of Carnap: "The prodigious growth of physics since the last century
depended essentially upon the possibility of referring to unobservable
entities like atoms and fields."43
It is interesting to compare this with Einstein's ideas about the role
of theoretical concepts which was formulated much earlier:
"These fundamental concepts and postulates, which cannot be further
reduced logically, form the essential part of a theory, which reason cannot
touch. It is the grand object of all theory to make these irreducible elements
as. simple and as few in number as possible, without having to renounce the
adequate representation of any empirical content whatever. "44
If one interprets the slightly unclear words "the essential part of a
theory, which reason cannot touch" to mean "not to be derived from
experiment, but nevertheless endowed with an important theoretical func-
tion",. one concludes that Einstein has anticipated the long process of
semantical liberation in his scientific practice as well as in his methodological
reflections - although without the necessary steps of justification.
One of the central questions in the discussion about the meaning of
theoretical entities was the relation between these abstract terms and real
entities. It was again Carnap who introduced an important distinction into

41 Cb. Misner et al.: Gravitation, San Francisco, 1973, p. 400.

42 R. Carnap: The Methodological Character of Theoretical Concepts, Minneapolis,
1956.
43 R. Carnap: Intellectual Autobiography, I. C., p. 80.
44 A. Einstein: On the Methods of Theoretical Physics, in: idem: Essays in Science,
New York, 1934, p. 15.
152 Bernulf Kanitscheider

the discussion. This distinction concerns non-empirical entities contained

in factual theories (for instance, elementary particles, fields) as well as the
abstract objects of formalistic sciences (for instance, classes, propositions,
numbers). According to Carnap one has to distinguish between two types
of existence problems: problems connected with the introduction of a
certain linguistic frame-work (called external problems) and those which can
be formulated only within, or to be more precise, after the admission of a
certain linguistic form (internal problems).45 Ontological claims can be
validated only within a certain linguistic frame while decisions about the in-
troduction of this linguistic frame-work itself can be decided only with the help
of pragmatic arguments. The question whether a black hole exists in Cygnus
X-I, for example, is a meaningful ontological question since it refers to an
element of the region of denotation of a linguistic frame which has been
introduced before. In contrast to this no answer can be given to the question
concerning the existence of things in themselves. This is a meta.physical
problem and posed incorrectly. One can find arguments for the introduction
of a realistic linguistic frame-work, but these arguments are of a pragmatic and
not of cognitive nature. Theoretical knowledge can influence the choice of
an object language (reism), but it would be incorrect to make use of the
efficiency and simplicity and fruitfulness of a realistic language in everyday
life and in science for a decision of the external problem concerning the
factual truth of the frame itself. In accordance with Carnap's tolerance for
linguistic forms but also their undecidability, Quine formulated a clear
criterion which helps to determine the presuppositions about existence
made in a theory. If we express the statements of a theory in canonical
notation with the help of quantifiers then the set of all objects, the existence
of which we assume, are the region of definition of the bound variables.
Ontological problems can be solved in the end only by decision, and state-
ments about existence have cognitive meaning only in well limited regions:
"Such talk of subordinate theories and their ontologies is meaningful, but
only relative to the background theory with its own primitively adopted and
ultimately inscrutable ontology." 46
After the edict of tolerance by Carnap and after Quine's release of
ontological decisions it was possible for a scientist to assign a reference of
this relative kind to theoretical terms which goes much beyond the visible.
In an important paper of 1950 47 , H. Feigl pointed out the surplus
meaning which a theoretical concept has in the set of objects which are not
empirically accessible. It became clear that a theory has a larger factual

45 R. Carnap: Empiricism, Semantics and Ontology, Revue Int. Phil. (Brussels) 4, 11

(1950), p. 20-40.
46 W. V. O. Quine: Ontological Relativity, New York, 1969, p. 51.
47 H. Feigi: Existential Hypothesis, Phil. Sci. 17 (1950), p. 35-62.
Einstein's Treatment of Theoretical Concepts 153

reference and may refer to objects of a completely different type than those
given by the set of test instances (empirical evidence). This agrees with the
fundamental postulate of hypothetical realism mentioned in the first section.
A classical example for this are the statements about our past in which the
reference object is inaccessible for nomological reasons (the world in which
we live contains no causal loops). A glaciological hypothesis, for instance,
contains the statement that the last alpine glacial period took place about
9,000 years ago. The statement contained in this hypothesis refers to an
object which no longer exists (the high level of glaciers at that time),
but can be checked by a finite set of observations concerning glacial
polishing and moraine deposits. A causal connection has to be assumed,
of course, between the traces we find from the past and the past events
themselves. In our example this is the dynamics of the creation or moraines,
but this can be tested in the present time because we can observe directly
from presently advancing glaciers the displacement of rock as well as the
process of glacial polishing.
Even beyond the realm of statements concerning the past it has turned
out to be meaningful to distinguish clearly between the relation of reference
and the relation of evidence. In this respect Einstein's theory of gravitation
was epoch-making. In its cosmological applications it contains statements
about large trans-empirical space-time regions but can be tested only in a
few local places. The very success of such a theory makes apparent how
decisively the positivistic command is violated to include only sense data
into the reference class of scientific terms. No excessive use of the surplus
meaning must be made, however. A mathematical formalism can develop
a certain internal life for which not all consequences can be interpreted as
really existing. This is certainly the case with the infinitely many worlds
which come up in the sequence of the analytic extension of the Reissner-
Nordstn6m and the Kerr metric 48 It is no simple task to find the limits be-
tween the surplus meaning and the runaway ontology of a formalism. Non-
empirical criteria of a philosophical nature will be essential in solving this task.
Apart from this difficulty it is not surprising that Einstein's epistemological po-
sition turned towards the hypothetical realism (which had corresponded to his
scientific practice from the beginning) when he constructed his general
relativity which contains a great wealth of theoretical concepts (metric,
affinities, curvature expressions).49 The factual necessity which led to an
increased use of epistemologically abstract objects in the construction of
general relativity has certainly influenced Einstein's metatheoretical reflec-
tions as can be seen in his statement concerning Mach (1922):

48 Cf. S. W Hawking/G. F R Ellis: The Large-Scale Structure of Spacetime, Cambridge,

1973, p. 149 ff.
49 G. Holton: Einstein, Mach and the Search for Reality, Daedalus 97 (1968), p. 636-
673.
154 Bernulf Kanitscheider

"Mach's system studies the eXIstmg relations between data of ex-

perience; for Mach, science is the totality of these relations. That point of
view is wrong, and, in fact, what Mach has done is to make a catalogue,
not a system; "50
Einstein is distinctly opposed to Mach's descriptivism. It is not the search
for fundamental nomological structures that is the epistemological goal of this
descriptivism, but the universal state matrix, the complete enumeration of
events from which the patterns of laws governing the course of these events,
however, cannot be derived. The reason for the ideal of a complete and eco-
nomic description prevailing at that time was the restriction to observational
concepts considered to be necessary. Theoretical elements are contained,
however, in almost every law with the exception of the lowest ranking
empirical generalization. It is remarkable that at a time when the philosophy
of science tried very hard to eliminate all theoretical entities, Einstein had
already recognized the unavoidable status: "Over against that I see no
'metaphysical' danger in taking the thing (the object in the sense of physics)
as an independent concept into the system together with the proper spatio-
temporal structure."51
It is obvious that Einstein's arguments are not of the refined form of
epistemological deductions, which philosophers gave later, but it is remark-
able that his intuition had led him along the correct way.
It became clear relatively late that the procedures introduced by
Ramsay52 and Craig 53 could not be used to bring about an ontological
decision concerning the existence of unobservable entities, even if theoretical
terms could be eliminated with certain restrictions. 54 This can be seen if
one accepts the dubious separation of the vocabulary V of a theory T into
a theoretical part VT and an observational part Vo. Denoting the set of
axioms of T by A, one may regard the conjunction of axioms of A to be a
statement of the form F (<fI I , <fI 2 , .•• , <fin) where <fI are the theoretical
predicates of A belonging to V T. If one wants to avoid the quantifica-
tion over predicates, one can rewrite a sentence in a set theoretic lan-

50 A. Einstein: Theory of the affine field, Nature 112 (1923), p. 448-449. Cf. also his
contribution to "Theorie de la relativite". Soc. fran~. phil Bull. 22 (1923), p. 111 ff.
51 A. Einstein: Bermerkungen zu Bertrand Russels Erkenntnis-Theorie, in: P. A. Schilpp
(ed.): The Philosophy of Bertrand Russell, La Salle (Ill.), 1971, p. 29l.
52 F. P. Ramsay: Foundations of Mathematics, 2nd edition, London, 1978, p. 233-
241.
53 W. Craig: Replacement of Auxiliary Expressions, Phil. Rev. 65 (1956), p. 38-55.
54 C. G. Hempel: The Theoretician's Dilemma, Minnesota Stud. Phil. Sci., vol. II,
Minneapolis, 1958, p. 37-98; G. Maxwell: The Ontological Status of Theoretical
Entities, Minnesota Stud. Phil. Sci., vol. III, Minneapolis, 1962, p. 3-27;].]. C.
Smart: Between Science and Philosophy, New York, 1968, p. 145. (We follow here
Smart's line of thought.)
Einstein's Treatment of Theoretical Concepts 155

guage as F' (K] , K 2' ... , Kn) where Ki is a class of m-tuples for which
the m-valued predicate <Pi is true. The Ramsay-procedure consists in replacing
F'(K],K 2 , ... ,Kn ) by3X!,3X2 ... 3Xn F'(X!,X 2 •.• X n ) where the
constants K! ... Kn have been eliminated with the help of bound variables.
In a Ramsay-sentence of this type belonging to T, one does not use, for
example, the predicate <PK = "is an electron" or Kk = "the class of electrons",
but only the expression "there is an Xk such that ... ". Thereby V T is
eliminated and a theory about electrons (for instance Dirac's theory) is
replaced by a theory about an abstractly denoted class of objects such that
the relations between the concrete elements and those between the abstract
classes are isomorphic. If the theory formulated in a realistic way originally
makes statements about the existence of electrons, the fact remains neverthe-
less, that the Ramsay-reformulation of the form " ... 3 Kk ... (3X) X E Kk"
contains the same ontological burden. Thus the Ramsay-reformulation can
be interpreted either in an instrumentalist way or with an existential claim
in the same way as the original theory: "The device of the Ramsay sentence
cuts no metaphysical ice one way or the other. "55
A similar ontological indifference applies to the Craig theorem which
replaces a theory T by another T' which contains no part of the vocabulary
of T but only deals with linguistic statements of T. Quite apart from the prac-
tical difficulty which arises from the fact that the empirical equivalence of T
and T' applies only if T' contains infinitely many axioms, there is the funda-
mental counter-argument that both methods of elimination can be applied
only to already existing theories. Only when theories containing theoretical
terms and references to unobservable entities have been created, can they
be converted with the help of one of the logical procedures into a phenom-
enological prediction machinery. Even if this procedure were practical in
all highly developed theories such as the theories of relativity, no questions
could be raised concerning the success of the corresponding replacement
theories T'. Nobody could answer the question why a certain black-box
connection T' of data is successful, since the phenomenalist must not use the
obvious answer of a realist who would argue that the success of the theory
is based on the fact that the objects referred to by T do exist.
Luckily Einstein did not have to follow all these detours aiming at the
elimination of theories and arrived in a rather direct way at the epistem-
ological conviction that no metaphysical harm is done by assigning a realist
interpretation to theoretical terms. 56
A final objection has to be considered: it was just this epistemological
position which led Einstein in his later years, despite all his efforts, into a
scientific back alley. His objections to quantum mechanics, which later con-

55 ].]. C.Smart,l.c .. p. 147.

56 A. Einstein: Physik und Realitat, Journal of the Franklin Institute 221, 3 (1936),
p. 313-347, esp § 1.
156 Bernulf Kanitscheider

cerned the completeness rather than the consistency of the theory, are
clearly connected with his criterion for reality formulated most distinctly in
his paper with Podolsky and Rosen. 57 Since quantum mechanics can be
interpreted only as a statistical theory concerning ensembles rather than a
complete theory of single events, and since he did not consider the introduc-
tion of hidden parameters, only the program of a unified field theory re-
mained, which cannot be realized according to the opinions of almost all
present theoreticians. Was Einstein's epistemological approach including its
ontological claim which he upheld until his death, a misleading guiding
principle? Unfortunately he did not witness the turn in the interpretation
of quantum mechanics which took place in the middle of the sixties. Philos-
ophers 58 and physicists 59 showed that quantum mechanics can be imbedded
in a realistic philosophy of science without any change in its formal struc-
ture and without any loss in its predictive power, but also without a change
in its fundamental probabilistic character. These ideas start from an objective
understanding of probability which guarantees that the revolution brought
about by quantum theory was a nomological revolution rather than an
epistemological one. The idealistic turn in the epistemology of quantum
mechanics which had been emphasized by the Copenhagen interpretation
and the orthodox interpretation of quantum mechanics and which had
been rejected by Einstein, is thus not connected with the statements con-
tained in the theory in an essential way.
In numerous analyses of Einstein's philosophy of science it has been
noted that his ideas were never sufficiently homogeneous and logically
coherent to be abbreviated with a well defined "ism". This applies even for
his explicitly trans-phenomenological period as shown by statements which
go beyond the semantic realism defined in the first section:
"Our experience hitherto justifies us in believing that nature is the
realization of the simplest conceivable mathematical ideas. I am convinced
that we can discover by means of purely mathematical constructions the
concepts and the laws connecting them with each other, which furnish the
key to the understanding of natural phenomena. Experience may suggest the
appropriate mathematical concepts, but they most certainly cannot be
deduced from it. Experience remains, of course, the sole criterion of the
physical utility of a mathematical construction. But the creative principle

57 A. Einstein/B. Podolsky/No Rosen: Can quantum-mechanical description of physical

reality be considered complete? Phys. Rev. 47 (1935), p. 777-780.
58 K. R. Popper: Quantum Mechanics without the observer, in: M. Bunge (ed.): Quantum
Mechanics and Reality, New York, 1967; M. Bunge: Philosophy of Physics, Dord-
recht, 1974.
59 G. Ludwig: Zur Deutung der Beobachtung in der Quantenmechanik, in: L. Kriiger
(ed.): Erkenntnisprobleme der Naturwissenschaft, Cologne, 1970, p. 428-434.
Einstein's Treatment of Theoretical Concepts 157

resides in mathematics. In a certain sense, therefore, I hold it true that pure

thought can grasp reality, as the ancients dreamed. "60
These words suggest a quasi-platonic ontology that can be found charac-
teristically also in Heisenberg's later works. It is essentially "quasi"-platonic
ontology, since neither Einstein nor Heisenberg wanted to forego an empirical
validation of the unified reconstruction of nature. The rationalism expressed
above concerns the conviction about the existence of a fundamentally
simple mathematical structure of the world which can be grasped by man
with sufficient creative formal fantasy. The quotation does not mirror an
apriorism which vindicates the methodological independence of a certain class
of synthetic statements from postulates of testability, but it expresses the
ontological conviction that the fundamental patterns of the universe can be
reconstructed in the form of laws, despite its high degree of complexity.
It is obvious that the construction of a unified field theory or quantum-
field theory of matter presupposes strong ontological assumptions about
the nomological structure of nature.
While these rationalistic components of Einstein's epistemology find
their natural place within an analytical conception of science, especially
since no synthetical knowledge a priori is admitted, a different position has
to be taken with respect to another aim of his, in connection with the search
for unified field theory: Taking up Leibniz' ideal hetried not only to recon-
struct the fundamental law of nature mathematically, but also to understand
at the same time why nature has this and no other structure. 61 This aim is
occasionally connected with Einstein's religious convictions,62 but it can
also be interpreted in a secular way. If it were possible by some method to
find a theory encompassing all interactions, this fundamental dynamics
would be able to explain all natural processes in a hypothetical-deductive
way. No possibility would exist, however, to answer the question why
nature possesses this and no other fundamental structure - except if one is
satisfied with the "anthropic" answers presented by Dicke, Carter, and
Wheeler. According to this argument the cognitive system itself and therefore
the existence of men explains the nomological structure we find in nature. If
one is not willing to accept this idealistic turn of the direction of explanation
- and there are good reasons not to do S063 - one will conclude that no
complete reduction of the contingent elements of our knowledge is possible.
This reduction would have to prove that the boundary conditions and the

60 A. Einstein: On the Methods of Theoretical Physics, in: idem: Essays in Science,

New York, 1934.p.17.
61 A. Einstein: Ober den gegenwartigen Stand der Fe!dtheorie. Festschrift fiir Aure!
Stodola. Zurich, 1929, p. 126-132.
62 G. Holton: Mach, Einstein and the Search for Reality, I. c., p. 659.
63 B. Kanitscheider. Probleme und Grenzen eines naturalistischen Weltbildes. Un-
published manuscript.
158 Bernulf Kanitscheider

laws of our universe are logical necessities. Even if Einstein's highest scientific
aim remains inaccessible for epistemological reasons, the reduction of con-
tingency, the growth of nomological structures and the attempt to go as far
as possible back in the hierarchy of explanations remain a fruitful research
program.
At the moment the understanding for this high aim of scientific
knowledge is not present everywhere. There are many pragmatic constraints
like technical usefulness, economic use for the interests of society, or other
means of utilizations which are demanded from the results of science. The
purely cognitive aim to understand why nature is structured in such a
nomological way as we find it, is considered to be an esoteric epistemological
luxury not connected with the interests of human life. Einstein's epistemo-
logical position shows in retrospect that the highest intellectual accomplish-
ments are not achieved under conditions suffering from these pragmatic
restrictions and cognitive amputations. The pragmatic aims of technical
usefulness can be accomplished by science successfully only if they are not
used as motives in the search for the fundamental laws of nature.
 159

Einstein's Importance to Physics, Philosophy,

and Politics
Carl Friedrich v. Weizsacker

Imagine that several centuries from now there will be people who
are interested in phases of the history of man long since past, and let us
ask ourselves which single name in our century will have the best chance
of being known to them. Certainly, we contemporaries have been most
shaken by the political events. But our political crises and those responsible
for them will one day be overshadowed by the crises, and if we are lucky,
their solutions, which are yet to come. If solutions are to be found, perhaps
our radical politicians will appear too inhuman to the people of the future,
our humane politicians not radical enough; perhaps of all the great men of
our century, only Gandhi will stand up to their judgment. Our century's art
will perhaps be remembered as a seismograph of earth-shaking events. The
earth-quakes are triggered by technical progress, and this, in turn, is made
possible by science. But science is at its greatest, and is in the end most
effective, where it seeks not to change the world, but rather to find truth.
The most famous scientist of our century is Einstein.
Would we scientists among ourselves also acknowledge him as our re-
presentative? Do we consider his extraordinary fame as well deserved? As a
physicist he has a chance, because among the sciences, natural science is the
primary pillar of the new concept of the universe, and physics is the basic
discipline of natural science. Physics took two revolutionary steps at the
beginning of our century - the theory of relativity and the quantum t·heory.
One of these theories is Einstein's achievement; in the first phase of the
other theory, his contribution equalled that of Planck and Bohr. Perhaps
Einstein is an especially worthy representative of our guild, for the very
reason that he never really wholly belonged to it. To the world around
him he was a naive genius. Yet his naivete, the naturalness of his questions,
were the heart of his genius. He posed every question directly, certainly
not in contempt of the guild's knowledge, but never in the usual scheme of
questions. Others, too, could provide answers; he was a master of question-
ing. And an unconscious master at that; he could never ask other than
directly.
The editors of this volume assigned me this theme; I have changed
its formulation only slightly. It is not a matter of evaluating each work,
or recounting Einstein's life, but of grasping the way in which his works
160 Carl-Friedrich von Weizsiicker

partly form, partly reflect the intellectual movement of his century. With
this in mind I first want to avoid the distinction between physics and philos-
ophy. Einstein was a physicist and not a philosopher. But the naive direct-
ness of his questions was philosophical. In the strenuous yet smooth course
of the accumulation of knowledge through "normal science", philosophizing
is an obstacle, because, as Kuhn says, such science always adheres to a para-
digm, the not questioning of which is one of its conditions for success. How-
ever, in crises and their revolutionary solutions, the vital question is: "What
do we actually mean by the words we use?" And this - as we were taught
by Socrates - is the philosophical question. The reverse also applies. Not
only does natural science require being stirred up by philosophical question-
ing; philosophy also needs to be stirred up by scientific answers. I still re-
member from my youth how the theory of relativity challenged - and, for
those who could see, smashed - the apriorism of academic philosophy. Less
successful historically but just as significant is the aversion of the mature
Einstein to the empiristic theory of science, which has since triumphed. At
best philosophical realism has a right, albeit limited, to refer to Einstein.
We will encounter all of these questions in following Einstein's great scientif-
ic steps.
Einstein plucked the special theory of relativity like a ripe piece of
fruit. The question was derived from experience: the Michelson experiment.
The mathematical explanation of this experiment was presented by Lorentz
and Poincare. Einstein's contribution, technically speaking, was the simple
derivation of the Lorentz transformation from two general postulates. More
philosophically speaking, Einstein reaped fame for the discovery that this
transformation was not a complication but rather a simplification; that it was
not a problem but a solution to a problem which had previously not even
been seen as such by physicists.
That the relativity of motion is a problem was indeed understood
by the philosophers since antiquity. Aristotle artfully avoided the necessity
of introducing a concept of space by defining position and change of posi-
tion of an object only in relation to other objects; a method which is satis-
factory for a closed, geocentric universe. When the theory of the infinite
universe was taken up again in modern times the problem arose again. New-
ton invented the metaphysical entity of absolute space as a solution. It is
remarkable with what respect Einstein speaks of Newton's proposal. Einstein
was one of that rare breed of revolutionaries, who really change something;
they have this capability because they have grasped the inherent logic of that
very preceding solution which it is their fate to surmount. Einstein did not
simply follow the idea of mechanical relativity, which Leibniz and Mach had
advanced against Newton. He remained faithful to the physical problem of
his time. In optics and electrodynamics physicists postulated the existence
of a physical substance, ether, which made possible the re-establishment of
the Aristotelian solution (hardly known to them any more), accounting
Einstein's Importance to Physics, Philosophy, and Politics 161

for the problem of the relativity of motion even in an infinite universe.

Maybe mechanical motion actually meant motion against ether? The Michel-
son experiment destroyed this solution, and hence this experiment appeared
to be a calamity. Einstein began by fully accepting this result, rejecting
the idea of an ether as an unnecessary solution to a non-existing problem
("Scheinproblem"), and returning to the relativity of motion. But the
experience of local action, more precisely, the invariance of the speed of
light, presented him with another problem that no one, with the exception
of Kant, had even considered: the concept of the relativity of simultaneity.
Kant! explained the simultaneity of distant events by interaction, which he
understood to be a force at a distance. Einstein (who could hardly have
known this passage of Kant's) solved the same problem in the opposite way
in accordance with the knowledge of his times: since action progresses with
finite velocity, simultaneity of distant events is not necessarily an objective
fact. Then its relativity contained in the Lorentz transformation is no longer
a calamity but a solution to a problem which had not even been seen be-
fore.
What has happened here epistemologically? Einstein put a demand
on the theory, which we can call the demand of semantic consistency.
Heisenberg had assumed in a very memorable conversation 2 with Einstein
in 1925 that Einstein had postulated the limitation of the theory to observ-
able quantities as he himself had done in his recent work on quantum
mechanics. Einstein corrected him by saying, "Only the theory determines
what is observable." The theory has so to interpret those physical experi-
ences which give a physical meaning to its mathematical concepts that the
interpreted phenomena will obey the laws of the very theory (provided
the theory is correct), only under this condition is it possible to empirically
test the theory. In this regard, Einstein speaks of theoretical concepts in
his later years as free human creations, for example: "The system of con-
cepts is a creation of man together with the rules of syntax, which consti-
tute the structure of the conceptual systems. Although the conceptual
systems are logically entirely arbitrary, they are bound by the aim to permit
the most nearly possible certain (intuitive) and complete co-ordination with
the totality of sense-experiences; secondly they aim at greatest possible
sparsity of their logically independent elements (basic concepts and axioms),
i.e. undefined concepts and underived (postulated) propositions." 3
These sentences summarize with direct simplicity the methodological
experience of a scientist's lifetime and set this experience apart from the
opinions of all other schools of philosophers. Einstein's view is not apriorism:

Critique of Pure Reason, A 211 f., B 256 f.

2 W. Heisenberg, Der Teil und das Ganze, Munich, 1969, p. 92.
3 P. A. Schilpp (ed.), Albert Einstein: Philosopher-Scientist, Evanston, Ill., 1949,
p. 13.
162 Carl-Friedrich von Weizsiicker

the conceptual systems are in themselves "entirely arbitrary". His view is

not empiricism in the traditional sense: the correspondence to experience
is, to be sure, a criterion a posteriori, but it is not the origin of the concepts.
It is not realism in the usual sense, since it does not presuppose a "real
world", but only "the totality of the sense-experiences"; I will return later
to its relationship to realism. It is not conventionalism: the conceptual sys-
tems are "arbitrary, but bound by the aim ... ". That which in our terms
can be recognized as conventional Einstein describes by his term of rela-
tivity. The Special and General Theories of Relativity are for him the de-
scription of the objective laws which rule the transformation from one of the
various legitimate ways of describing the occurences into the others. Ein-
stein does not either replace all these philosophies by a new one. He remains
a physicist who says what he has experienced and leaves open what he does
not know. Neither does he show contempt for philosophy; philosophical
questioning is essential to him.
The General Theory of Relativity is Einstein's most individual achieve-
ment. Of all the well-known theories of physics, it is the only theory where
one may doubt that it would have already been discovered at all had Einstein
not lived. This may be said regardless of the simultaneous independent
discovery of the basic equations by Einstein and Hilbert 4 , for the formula-
tion of the questions in the years preceding the discovery was Einstein's
work. For exactly this reason, the true relation of this theory to the rest of
physics has not become really clear to this day. The theory is like a down-
payment on something yet unknown which has not been paid out; this is
the way Einstein himself felt about it.
This is reflected in its philosophical problems which are still not solved
today. This pertains not to those features which shocked philosophers so at
the outset, such as the curvature of space, the connection of geometry with
experience, and more specifically, of geometry with dynamics. With this
Einstein upheld the geometric tradition of the 19th century, especially the
thoughts of Riemann. Einstein gave Riemann's ideas a certain resonance
that made even the philosophers sit up and take notice. His own contribu-
tion lay in the connection of two other thoughts - the philosophically
abstract one of general relativity and the physically concrete one of the
principle of equivalence. He needed Riemannian geometry in order to
connect these. The principle of equivalence was a trick of genius in the
process of naive, direct questioning. If two quantities - here inertial and
gravitational mass - are empirically always equal, there ought to be some
theory which states they are identical. But with general relativity the unan-
swered questions begin. Einstein's philosophical instict told him that this
was something of essence, and it resulted in the formulation of the heuristic

4 cf]. Mehra, Einstein, Hilbert, and the theory of gravitation, in ]. Mehra (ed.),
The Physicist's Conception of Nature, Dordrecht, Reidel, 1973.
Einstein's Importance to Physics, Philosophy, and Politics 163

principle of the general covariance of basic equations. He was yet to learn

that this principle can always be satisfied, and he reworded it to state that
the correct equations must be especially simple to derive in general covariant
formulation.
But what does simplicity mean here, and how is it justified? Einstein
thought the principle to be very meaningful epistemologically, particularly
in regard to the criticism of empiricism. "I have learned something else from
the theory of gravitation: No ever so inclusive collection of empirical facts
can ever lead to the setting up of such complicated equations. A theory can
be tested by experience, but there is no way from experience to the setting
up of a theory. Equations of such complexity as are the equations of the
gravitational field can be found only through the discovery of a logically
simple mathematical condition which determines the equations completely
or (at least) almost completely. Once one has those sufficiently strong for-
mal conditions, one requires only little knowledge of facts for the setting up
of a theory; in the case of the equations of gravitation it is the four-dimen-
sionality and the symmetric tensor as expression for the structure of space
which, together with the invariance concerning the continuous transforma-
tion-group, determine the equations almost completely. "s
A modern physicist who reads these lines may tend even to intensify
the example. The four-dimensionality and use of the tensor concept appear
to be the "little knowledge of facts" which Einstein refers to; the remainder
is Riemannian geometry. What Einstein is actually introducing here is the
local invariance under the Lorentz-group (the non-homogeneous, now
known as the Poincare-group). So Einstein makes use of two groups at the
same. time: Poincare-group as a local group and the continuous group corre-
sponding to it as a global group. It remains one of the unsolved tasks of phys-
ics to account for this. Furthermore, there are still infinitely many possible
field equations, that satisfy Einstein's claims; of these, Einstein's are unique-
ly determined as the simplest ones. In this postulate of uniqueness he was so
serious that he later radically rejected his own additional "cosmological"
term. And finally, Einstein knew from the very beginning that a theory
not involving space alone, but space and matter together, had to be found.
He could evade the unsolved problem of matter by means of the postulate
of invariance: in an approximate theory matter was to appear only in the
disguise of its energy tensor. When considering all the questions which have
remained unsolved in the six decades since 1915, it may seem miraculous
that Einstein succeeded in setting up a simple and empirically correct theory
of gravitation. Under a personal aspect we may see this as a token of his
instinct for simplicity, of his genius. Philosophically, however, one must
admit that physics since then has not even been able to formulate clearly

5 Schilpp, l.e., p. 89.

164 Carl-Friedrich von Weizsacker

the essence of the questions which are waiting for us - unanswered. What
we are missing here is - as it was the case before the special theory of rela-
tivity - not the answer but the simplicity of the question.
One has to admit that in the last forty years of his life, Einstein himself
was not successful at continuing this simplicity of questioning. This, even
more than the horrible political events, was the tragic shadow that darkened
the second half of his life. But we can learn just from the attempt at under-
standing him in this period.
Einstein formulated the problem of the theory of space and matter
as the problem of a uniform field theory. He criticized Newton for viewing
space as a "tenement building" in which bodies move in and out like tenants.
Mach wanted to accept the concept of matter as an economical description
of sensations, and reject the concept of space as a metaphysical invention.
Einstein, who began under Mach's influence, was led to the opposite solution
of the dualism of space and matter. His claim of simplicity under general
covariance forced a non-linear field equation of gravitation upon him. Here
he also hoped to find the simple solution to a problem of which nobody had
been aware before, by facing an apparent calamity and complication. The
solutions to non-linear differential equations contain singularities. These sin-
gularities might be the field lines of the mass-points. If so, then the task of
fundamental physics is reduced to designing a general field theory. Einstein
knew of only two fundamental fields - those of gravitation and electromag-
netism. After decades of unsuccessful attempts he came back to the older
thought of using a field-tensor of arbitrary symmetry as a basis and dividing
it into a symmetrical gravitation-tensor and a skew symmetrical electro-
magnetic tensor. However, he never succeeded in deriving a realistic picture
of matter.
Philosophically, Einstein proposed a conclusive theory of physics,
which, I presume, is historically due in our century or the next. Indeed, the
physics of the 19th century had already received him with the idea of such
a theory as he described ironically as an older man: "In the beginning (if
there was such a thing) God created Newton's laws of motion together with
the necessary masses and forces. This is all; everything beyond this follows
from the development of appropriate mathematical methods by means of
deduction."6 This vision of the world, as one of mass-point mechanics with
action at a distance, was in fact destroyed by the rise of field physics. Ein-
stein's conception was to make field physics a final uniform theory. But
what was the philosophical weakness of this world of mass-points? Not only
that the realities of the consciousness, the emotions and values were not
described in it. Einstein's design did not do that either. The weakness was
that one had to accept as given facts "as God created them in the beginning"
not only Newton's laws of motion, but values of masses and laws of force

6 Schilpp, I.e., p. 19.

Einstein's Importance to Physics, Philosophy, and Politics 165

as well. Einstein hoped, of course, that the field-theory would in principle be

capable of deriving all these data without introducing any arbitrary constants.
It was not at the heart of the tragedy that Einstein was in fact not capable
of accomplishing this. The mathematical complication of non-linear differen-
tial equations has defied many attempts at solutions since then, such as Hei-
senberg's late attempt at resuming a quantum-theoretical approach to a
uniform field theory. Our younger generation believes that the tragedy lay in
the fact that only a quantum theory, and not a classical field theory, had any
chance of solving the problem.
In the second half of Einstein's lifetime the quantum theory - under
the guidance of Bohr and his disciples - completed its victory. In the begin-
ning stages of this theory Einstein's vital accomplishment was to radicalize its
problems. Einstein's hypothesis of the quantum nature of light alone was
enough to deservedly earn him the Nobel Prize. Einstein had made an issue
of the dual nature of light. Yet he could not acknowledge Bohr's principle
of complementarity of waves and particles in a single statistical theory as a
legitimate solution. After dramatic debates with Bohr he had to admit that
the theory was free from contradiction, but he stuck to his decision: "The
Heisenberg-Bohr tranquilizing philosophy - or religion? - is so delicately
contrived that, for the time being, it provides a gentle pillow for the true
believer from which he cannot very easily be aroused."7 We younger physi-
cists were naturally convinced that Einstein's classical belief was just that -
a pillow to rest on. After a newspaper article concerning Einstein's revolu-
tionary political ideas had been read at Bohr's Institute, the comment was
heard: "Isn't it strange that Einstein is so conservative only in physics?"
Who was right~
Einstein wanted to have a classical, real field as a basis and hoped to
finally be able to explain particles with their quantum characteristics in
terms of classical, real singularities of this field. The quantum theorists
interpreted the Schrodinger field as a probability field of particles, viz.
not as factual, but as an expression of possibilities for characteristics of
particles. Thus they sacrificed at the same time the concept of orbit, i.e.
the classical concept of the particle. Empirical success was with quantum
theory. But we should give Einstein credit for maintaining that in the final
analysis, the problem must be solved through fundamental reasoning. In the
process Einstein sought the clearest formulation of the discrepancy in the
concept of reality of both sides. He found it in the famous thought experi-
ment of Einstein, Podolsky, and Rosen, in the inference drawn on the
condition of a system S2 by means of measurement of a system SI which
is spatially separated but was previously connected with it. This considera-
tion reveals the relevance of the quantum-theoretical concept of the reduc-

7 Letter to Schrodinger, May 1928. Quoted by B. Hoffmann and H. Dukas, Albert

Einstein. Creator and Rebel, Viking Press New York, 1973, p. 190.
166 Carl-Friedrich von Weizsiicker

tion of the Schrodinger wave by an act of measurement. The logical contra-

diction in the description of this experiment can be escaped from "only by
either assuming that the measurement of S 1 (telepathically) changes the real
situation of S2 or by denying independent real situations as such to things
which are spatially separated from each other. Both alternatives appear to me
entirely unacceptable."8 The solution of this apparent paradox according to
quantum theory is indeed Einstein's second "inacceptable" alternative: spa-
tially separated objects have not objective independent "real situations" in
as far as the observational situation excludes this. That some quantum theo-
rists also perceive this as an apparent paradox, probably stems from the fact
that today's quantum field theory treats spatial distances as classical quanti-
ties. A description consistent with the quantum theory would presumable
have to view the whole metric field and all the space-time coordinates des-
cribed by it as a construction just existing as a classical limit. This, however,
has not yet been carried out.
But what is the philosophy behind Einstein's statement that the solu-
tion as given by quantum theory is "inacceptable"? It cannot be derived
from his previously cited thoughts on the physical conceptual system; they
could have been Heisenberg'S own thoughts in his younger years. Einstein's
"realism" which goes beyond them is expressed in the following lines:
"Physics is an attempt conceptually to grasp reality as it is thought inde-
pendently of its being observed. In this sense one speaks of 'physical reali-
ty'."9 Einstein saw his opponents not only as "empiricists" who, besides
rightly testing laws by experience, wanted to derive them solely from expe-
rience (an endeavor which is totally in vain). He also saw them as "posi-
tivists" who recognized only those things as real which could be observed.
Here he may have described a prevalent prejudice - or perhaps only a
prevalent convenient way of thinking - correctly. But he certainly did his
most significant scientific opponent (and personal friend)· Bohr an injustice.
Bohr's reasoning centered precisely around the question under what condi-
tions we can describe the perceived by way of an "objective", "unambigu-
ous" model of events; he was only convinced about the limitations of this
possibility. I assume that these questions will only be able to be fully under-
stood philosophically by an analysis of the essentially temporal character of
all cognition. Einstein's concept of reality is oriented around the phenome-
non of factuality, i. e. the past. The quantum theory's concept of probability
is oriented around the phenomenon of possibilities, i.e. the future.
Einstein's decision in this conflict was metaphysically determined, and
he knew it. In conversation he occasionally proposed a philosophical argu-
ment by apparently playfully referring to God: "God does not play at dice",
or "The Lord is clever, but he is not malicious." When put to the question
he answered directly: "I believe in Spinoza's God who reveals himself in
8 Schilpp, I. C., p. 85.
9 Schilpp, I.c., p. 81.
Einstein's Importance to Physics, Philosophy, and Politics 167

the orderly harmony of what exists, not in a God who concerns himself
with the fates and actions of human beings."10
It was particularly temporality which he felt to be only subjective.
"It seems to me that the idea of a personal God is an anthropological con-
cept which I cannot take seriously. I feel also not able to imagine some
will or goal outside the human sphere. My views are near those of Spinoza:
admiration for the beauty of and belief in the logical simplicity of the order
and harmony which we can grasp humbly and only imperfectly. I believe
that we have to content ourselves with our imperfect knowledge and under-
standing and treat values and moral obligations as a purely human problem -
the most important of all human problems. "11 It is justified to say that he
took a stance consciously outside of his Jewish religious tradition, as Spino-
za himself, but within the realm of European metaphysics which stemmed
from Greek thought. If the mode of reasoning was Greek, the particular
moral earnestness with which Spinoza and Einstein applied them was deeply
Jewish. In this frame of thought Einstein was able not only to think but to
live a full life as well. Four weeks previous to his own death he wrote to the
surviving family members of his friend Besso: "And now he has preceded me
briefly in bidding farewell to this strange world. This signifies nothing. For
us believing physicists the distinction between past, present, and future is
only an illusion, even if a stubborn one. "12
Whereas Einstein's philosophy is inseparable in content from his phys-
ics, his political influence is connected with his scientific thought only by
his unmistakable personality and by the consequences of his fame, which he
could sometimes hardly bear. He commented openly about his aloofness
from society: "My passionate sense of social justice and social responsibility
has always contrasted oddly with my pronounced lack of need for direct
contact with other human beings and human communities. I am truly a 'lone
traveller' and have never belonged to my country, my home, my friends, or
even my immediate family with my whole heart. In the face of all these ties
I have never lost a sense of distance and a need for solitude - feelings that
increase with the years."!3 He wrote this in 1930. And later, words of thanks
to Hermann Broch for his Vergil: "I am fascinated by your Vergil - and
am steadfastly resisting him. The book shows me clearly what I fled from
when I sold myself body and soul to Science - the flight from the I and
WE to the IT."14
It was just this aloofness which enabled Einstein to remain true to his
naive directness in his political judgment. In this respect he is more deeply
bound to the great political tasks of the future than many of us, who have

10 Hoffmann and Dukas, I.e., p. 95.

11 Hoffmann and Dukas. I. c., p. 95.
12 Hoffmann and Dukas, I. c., p. 257 f.
13 Hoffmann and Dukas, I.e., p. 253.
14 Hoffmann and Dukas, I.e., p. 254.
168 Carl-Friedrich von Weizsacker

conceded - apparently or truly - to believe in the political prejudices of

the present in order to be able to function concretely in their midst.
Biographically it is necessary to take a look at his fourfold nationali-
ty: German, Swiss, American, and Jewish. Even as a schoolboy in Germany
he suffered under the traditional authoritarian manner of thought and be-
havior. In spite of the hard times of his younger years he was much happier
in the liberal ambience of Switzerland. After the First World War, however,
he was ready to use his fame to act as a go-between for his German father-
land and its opponents. The factor in determining his loyalties was always
compassion with the suffering. He declared solidarity with the Germans
in 1919 when they had become the "world's enemy". He united with the
Jews after having experienced growing anti-Semitism. Not because of the
personal attacks, as much as these hurt him, but because of the inhumanity
of the sentiments against the Jews, which later resulted in inhumanity of
actions, was he repelled by Germany; he never forgave Germany for this.
Academic America offered him his last peaceful homeland, in complete
liberality. He criticized political America painfully and in vain when it
took the path to imperialism which is prescribed for all super-powers; he
was as foreign there as everywhere in the world of power politics.
His early political convictions showed him to be a pacifist - the only
natural stand for a person to take who is so independent of society. But his
pacifism was not a fixed doctrine; it was a naively direct judgment on the
consequences of the competition for power. Thus he was even able to shock
the pacifists by declaring the armed resistance to Hitler necessary. Many
people blamed him for signing the letter drafted by Szilard which initiated
the development of the atom bomb - and how could he possibly have been
happy about the results? It was his destiny to be asked for help in many
things which he could not quite get to the bottom of. But at the decisive
times it was he himself who reacted directly and made the decisions, not
those who had come to him for advice. He saw a commitment to reason
in Szilard's argument - should Hitler alone have this weapon? - and he
submitted to the logic of the historical process in which he found himself
entangled. It was actually his suffering due to politics which made his
political voice credible. Shortly before his death he signed the manifest
initiated by Bertrand Russell which instigated the Pugwash-Movement.
Seen politically, this movement became a not entirely insignificant instru-
ment of inofficial diplomacy, a forum for discussion in which scientists,
relieved of official political duty, could carryon preliminary discussions
on subjects later to be taken over by the politicians. Morally, the movement
was even more important, even if less effective in a direct way. It was one
of the places where truths - otherwise suppressed in the conflict for pow-
er - became audible. For in the struggle for power, truth cannot be spo-
ken, and yet, in the end, truth decides the historical power struggle, even
where truth and the outcome are tragic.
 169

Einstein and Zionism

Banesh Hoffmann

In 1911, at the age of thirty-four, Einstein faced a dilemma. Having

been offered a professorship at the German University in Prague, he was
required to fill out an official form. One of the questions asked about his
religious affiliation. It seemed, at the time, a simple question to answer.
Coming from Jewish parents, even though they were not observant of the
strict rituals of Judaism, Einstein well knew that he was a Jew. But since
he resolutely did not belong to any official Jewish or other religious group,
he answered the question on the form with the word "unaffiliated."
Soon, however, he learned that this response could have serious con-
sequences. To become a professor at the German University in Prague, he
would have to swear allegience to the Emperor Franz Josef, and if a candi-
date did not declare a belief in an officially recognized God by which to
swear, he could not swear an oath of allegiance that the Emperor would find
acceptable. The much-desired professorship was thus in jeopardy.
The problem, though serious, was not beyond Einstein's capacity for
ingenious solution. He asked the official in charge of the records to change
the entry about his religious affiliation. This the official refused to do with-
out further proof. Einstein then asked on what basis the official listed him
as "unaffiliated," and the official replied, probably heatedly, that it was on
the basis of Einstein's own declaration. The official must have felt that that
was the end of the matter. But Einstein responded by. saying that he, Ein-
stein, now solemnly declared himself to be of the Hebrew faith. And since
there was no logical alternative left open to the official, he changed "unaffil-
iated" to "Mosaic," which was the official term for one of the Jewish faith.
Actually "Mosaic" was a misnomer. Einstein was, indeed, a profoundly
religious man, but, as he repeatedly said, he did not believe in a personal
God. Rather he believed in the immanent God of the philosopher Spinoza,
and it is not irrelevant to mention that although Spinoza was a Jew, he had
been excommunicated by his Jewish confreres. Here are Einstein's own
words in response to a letter in 1947 asking about his views on belief in a
Supreme Being:
"It seems to me that the idea of a personal God is an anthropological
concept which I cannot take seriously. I feel also not able to imagine some
will or goal outside the human sphere. My views are near those of Spinoza:
admiration for the beauty of and belief in the logical simplicity of the order
B. Z. Mosessohn, A. Einstein, C. Weizmann and M. Ussishkin on board of the "Rotterdam"
travelling to the United States in 1921.
(Bildarchiv Preuflischer Kulturbesitz, Berlin, Germany)

and harmony which we can grasp humbly and only imperfectly. I believe
that we have to content ourselves with our imperfect knowledge and under-
standing and treat values and moral obligations as a purely human problem -
the most important of all human problems."
As a child, Einstein was sent, with his sister Maja, to the local Catholic
elementary school, and later to the Luitpold Gymnasium, where religious
instruction in one's own faith was compulsory, as was attendance at services.
The children also learned about Judaism at home from a private instructor.
Meanwhile, experience was teaching them basic facts about being
Jewish, as can be seen from the following excerpt from a draft of a letter
written by Einstein in 1920. He was some forty years old when he wrote it,
and at the height of his fame. Note how indelibly certain childhood happen-
ings had impressed themselves on him:
"The teaching staff of the elementary school was liberal and made no
denominational distinctions. Among the Gymnasium teachers there were a
Einstein and Zionism 171

few anti-Semites, one in particular who never let us forget that he was a
reserve officer. Anti-Semitism was evident among the children, particularly
in the elementary school '" Physical assaults and insults were frequent on
the way to school, though for the most part not really malicious. Even so,
however, they were enough to confirm, even in a child of my age, a vivid
feeling of not belonging."
The young Einstein was so impressed by his Jewish religious instruction
that he quickly became intensely religious in a formal, ritualistic sense,
refusing to eat pork and regretting that his parents did not observe the
Jewish rituals. This phase did not last. As he wrote at age sixty-seven in
his Autobiographical Notes:
"This deep religiosity came to an abrupt end at age twelve. Through
the reading of popular scientific books I soon reached the conviction that
much in the stories in the Bible could not be true. The consequence was a
positively fanatical orgy of freethinking coupled with the impression that
youth is intentionally deceived by the state through lies. It was a crushing
impression. Suspicion of every kind of authority grew out of this experience
... an attitude that has never left me '" "
It is perhaps worth mentioning here that some forty years after this
crushing experience Einstein had the grace to say, "To punish me for my
contempt for authority, Fate made me an authority myself."
For many years after his break with the Bible, Einstein seems to have
been little concerned with Judaism. While he was at the German University
in Prague he did come in contact with Zionists, but apparently he remained
quite without interest in Zionism. From Prague he was quickly invited to
the Zurich Polytechnic, where he had been a student - and where he had
initially failed the entrance examination. And from there in 1914, as is
well known, he went to a prestigious position in Berlin. But still Jewish
matters do not seem to have occupied his mind to any significant extent.
An indication of what happened next can be gleaned from the follow-
ing two quotations, even though they slightly conflict with previously quot-
ed material. Both were written in 192 9. In an article, Einstein said, "When
I came to Germany fifteen years ago I discovered for the first time that I
was a Jew, and lowe this discovery more to Gentiles than to Jews." And in
a letter he wrote, "1 first came to Zionism after my emigration to Berlin in
1914 at the age of 35, after I had lived in a completely neutral environment."
Let us not jump to conclusions. Einstein did not become a Zionist in
1914. Re-read his words: "1 first came to Zionism after my emigration to
Berlin at the age of 35."
It was actually quite a while after. As Kurt Blumenfeld said, "Till
1919 Einstein had no connection with Zionism or Zionistic ideas." As for
Kurt Blumenfeld, he was the director of propaganda for the Zionist Union
of Germany [Zionistische Vereinigung fur Deutschland]. The German
Zionists had made a list of Jewish intellectuals whom they wanted to attract
172 Banesh Hoffmann

to the Zionist cause, and Einstein's name was on the list. And so Blumen-
feld presented himself to Einstein in the February of 1919 to talk with him
about Zionism.
Note the circumstances. Two years earlier, the famous Balfour Decla-
ration had promised a national homeland for the Jews in what was then
called Palestine. While Zionists rejoiced, many successful assimilated Jews
were bitterly outspoken in their opposition, fearing that a Jewish national
homeland might tend to give them the status of aliens in the lands where
they had achieved their success. As for Einstein, he had propounded his
general theory of relativity in 1915, yet in early 1919 he was still relatively
unknown to the general public. The results of the British eclipse expedition
under Eddington were not officially announced till 6 November of that
year, and only then did the people of the world suddenly realize that there
was a mighty genius in their midst. When Blumenfeld first came to Einstein
in the February of 1919, Einstein's spectacular worldwide fame was some
nine months in the future.
Blumenfeld's task did not turn out to be easy. Einstein was extra-
ordinarily naive about Zionism. As Blumenfeld remarked, "Einstein [warmed]
up to the Zionist idea only gradually and after long deliberation. He joined
the movement when he felt that it was actually a matter of a struggle for
spiritual freedom, for human rejuvenation, and when he became convinced
that the conquest of Erez Israel for the Jewish people was a conquest through
labor and that the movement was free from tendencies of profiteering and
exploitation. "
By 1920, Einstein was beginning to speak out in Zionistic tones, say-
ing, for example, "Only when we [Jews] have the courage to regard our-
selves as a nation, only when we have respect for ourselves, can we win the
respect of others."
This change of outlook had not been brought about solely by Blu-
menfeld. The end of World War I had brought a sharp rise in anti-Semitism,
and Einstein had seen for himself the shattering effect that it had had on the
refugees who fled to Germany from Eastern Europe.
On 10 March 1921, Chaim Weizmann, in England, sent a detailed tele-
gram to Blumenfeld. Weizmann was the leader of world Zionism - he was
to become the first President of the State of Israel. In his telegram he told
Blumenfeld to persuade Einstein to accompany Weizmann on a visit to the
United States to help raise funds for Zionist causes. In particular, Blumen-
feld was to stress to Einstein the need to raise money for the creation of a
Hebrew University in Jerusalem.
Blumenfeld used all his arts of persuasion, but Einstein would have
none of it. Right away he said, "I am no orator. I can add nothing convinc-
ing. You will only be using my name." And he defeated Blumenfeld's argu-
ments so easily that Blumenfeld gave up the battle. But as he was about
to take his leave, a last desperate thought came to him. He turned to Ein-
Einstein and Zionism 173

stein and said, "I do not believe that we can weigh arguments against each
other in this case. Our work can succeed when all of us are moved by a new
spirit of national discipline .... I do not know what [Dr. Weizmannl would
say to you in my place. But I know that he has been entrusted by the Jewish
people with the responsibility of realizing the Zionist program. Dr. Weizmann
not as an individual, but as president of the Zionist organization has ordered
me to persuade you to go to America, and I have the right to expect that
you subordinate your consideration to Dr. Weizmann's decision." To Blu-
menfeld's surprise, Einstein agreed. And that is how it came about that Ein-
stein and Weizmann travelled together to America on an ocean liner in the
cause of Zionism.
Weizrnann was a distinguished scientist. Telling about the boat trip
across the Atlantic, he said, "Einstein explained his theory to me every day,
and on my arrival I was fully convinced that he understood it."
Einstein was received in America with great enthusiasm. Many honors
were bestowed on him, and his presence on the platform turned out to be a
major asset. For his unguarded account of what happened, here are excerpts
from a letter that he sent from New York to his long-time friend Michele
Besso, whom he had thanked in the celebrated 1905 paper that set forth
the special theory of relativity:
"Two frightfully exhausting months now lie behind me, but I have the
great satisfaction of having been very useful to the cause of Zionism and of
having assured the foundation of the university .... I had to let myself be
exhibited like a prize ox, to speak an innumerable number of times at small
and large gatherings, and to give innumerable scientific lectures. It is a
wonder I was able to hold out. But now it is over, and there remains the
beautiful feeling of having done something truly good, and of having inter-
vened courageously on behalf of the Jewish cause, ignoring the protests of
Jews and non-J ews alike."
The journey to America enormously strengthened Einstein's commit-
ment to Zionism, and his sense of being a Jew took on a new profundity.
From now on, guided by Blumenfeld, he spoke and wrote frequently in sup-
port of Zionism. No longer did he object that the Zionists were using his
name. He realized that his fame was a unique asset entrusted to him by
Fate, and his conscience told him that he must let it be used in the cause of
Zionism, as in other causes he saw as worthy. It was, for him, a profound
moral obligation. Unlike too many other Jews in Germany, he felt a strong
bond of kinship with the bedraggled, poverty-stricken Jewish refugees from
Eastern Europe. He could see beyond their outward appearance, their air
of vagabondage, and their lingering fear, and he recognized them not just as
fellow human beings but also as fellow Jews who, therefore, had a special
claim on his sympathy. He did not run from them as if fearing contamina-
tion from their contact. He saw that, unlike some of the more assimilated
Jews in Germany, they retained a vivid sense of belonging - "a healthy
174 Banesh Hoffmann

national feeling ... not yet destroyed by the process of atomization and
dispersion. "
Here are excerpts from a speech he gave in England in 1921 on his
way home from America:
"When I moved to Berlin ... I realized the difficulties with which
many young Jews were confronted. I saw how, amid anti-Semitic surround-
ings, systematic study, and with it the road to a safe existence, was made
impossible for them. This refers specially to the Eastern-born Jews in Ger-
many .... These Eastern-born Jews are made the scapegoat of all the ills
of present-day German political life and all the after-effects of the war.
Incitement against these unfortunate fugitives, who have only just saved
themselves from the hell which Eastern Europe means for them today,
has become an effective political weapon, employed with success by every
demagogue. When the government contemplated the expulsion of these
Jews, I stood up for them, and pointed out in the Berliner Tageblatt the
inhumanity and the folly of such a measure. Together with some colleagues,
Jews and non-Jews, I started University courses for these Eastern-born Jews,
and I must add that in this matter we enjoyed official recognition and consi-
derable assistance from the Ministry of Education.
"These and similar happenings have awakened in me the Jewish national
sentiment. I am a national Jew in the sense that I demand the preservation of
the Jewish nationality as of every other. I look upon Jewish nationality as a
fact .... I regard the growth of Jewish self-assertion as being in the interests
of non-Jews as well as Jews. That was the main motive of my joining the
Zionist movement."
And later in the speech he says: "The main point is that Zionism must
tend to enhance the dignity and self-respect of the Jews of the Diaspora. I
have always been annoyed by the undignified assimilationist cravings and
strivings which I have observed in so many of my friends."
As Einstein now saw it, Zionism was a unique revitalizing and unifying
force for all Jews. Though steeped in tradition, it looked to the future.
A few months after his return from America he said, "For the last two
thousand years the common property of the Jewish people has consisted en-
tirely of its past .... Now all that is changed. History has set us a great
and noble task in the shape of active cooperation in the building up of
Palestine ... [which must 1 become a seat of modern intellectual life, a
spiritual center for the Jews of the whole world .... A Jewish University
in Jerusalem constitutes one of the most important aims of the Zionist
organization. "
In 1922, Einstein visited Japan, where he was greeted with extraordi-
nary enthusiasm. Early in 1923, on his way back, he visited Palestine, where,
too, he was greeted enthusiastically. But here his visit had an emotional
impact that could not possibly be duplicated in any other land. He was
pressed from all sides to settle in Jerusalem and wrote in his travel diary,
"The heart says yes, but the mind says no."
Einstein and Zionism 175

The highlight of his stay was the visit to Mount Scopus, the site on
which the Hebrew University was destined to be built. His presence there
brought to the Jews of Palestine a sense of joyous fulfillment, and the rev-
erence they felt was manifest in their introductory remarks inviting him to
speak from "the lectern that has waited for you for two thousands years."
In these few words are packed the tragedy and triumph of Jewry.
With the rise of Nazism in Germany, the position of the German Jews
became desperate. When Hitler seized power, Einstein was in California. He
never returned to Germany. Instead, he severed all official German ties and
spoke out against the Nazi tyranny with the fervor and fearlessness of the
ancient Hebrew prophets. In Princeton, where he settled in America, he
found ways to rescue friends and strangers from death at the hands of the
Nazis.
After World War II, with the Nazis defeated, he was invited to rejoin
various German organizations from which he had resigned. But he refused.
And his words of refusal reveal how profound had become his sense of
Jewish identity. For example, to one organization he said, "The Germans
slaughtered my Jewish brethren. I will have nothing further to do with
the Germans." And to another he said, "Because of the mass murder that
the Germans inflicted on the Jewish people, it is evident that any self-
respecting Jew could not possibly wish to be associated in any way with
any official German institution." He never relented.
When Weizmann died, in 1952, Einstein was invited to succeed him as
President of the State of Israel. He declined the invitation gently, citing his
lack of aptitude and experience for such a post, and adding. "I am the more
distressed over these circumstances because my relationship to the Jewish
people has become my strongest human bond, ever since I became aware of
our precarious situation among the nations of the world."
And in March of 1955, less than a month before he died, he wrote to
Blumenfeld saying, "I thank you, as I should have done much earlier, for
having helped me become aware of my Jewish soul."

This article is adapted from a speech entitled "Eistein and Zionism" delivered at the
Seventh International Conference on General Relativity and Gravitation, in Tel Aviv,
in June, 1974, and published in the Proceedings of that conference (Wiley, N.Y.,
and Israel Universities Press, Jerusalem, 1975). I am grateful to Dr. Otto Nathan for
permission to quote material belonging to the Estate of Albert Einstein; and to Helen
Dukas for invaluable help in providing copies of documents from the Einstein Archives
and sharing with me her unrivalled knowledge of Einsteiniana.
 177

Birth and Role of the GRG-Organization and the

Cultivation of International Relations
among Scientists in the Field
Andre Mercier

Scientists in our days, and among them physicists, especially theoretical

physicists, ressemble monks from the Middle Ages in at least two respects:
They are at the front of learnedness, and they travel all the time long dis-
tances in order to exchange ideas. They even sometimes quarrel about points
of divergence, but science in our days has a greater degree of certainty than
philosophy in old days. Learned monks of old were mainly concerned with
proofs of God; theoretical physicists are, for a good number, concerned with
proofs of theorems within the new theories of the world. Cosmology has
shifted from a theological to a scientific discipline.
The Church was in the Middle Ages the only possible frame for interna-
tional organization of learnedness. At the dawn of Modern times, the Church
lost gradually its prerogatives as of giving all the necessary instructions for
both the praxis of daily life and behaviour in matters of the spirit. The
State gradually took over the role of telling everybody what they ought to
do until it became that huge conception of Hegel's and that monstrous
organization realized in ever so many States of capitalistic, communist or
other ideological structure and function. But the State did neither take
over the role of being a frame for world wide communication in spiritual
matters, nor did it produce the men who promoted further the advancement
of knowledge, for on the one hand States became more and more national,
and their intercourse became an ever more delicate matter of diplomacy
and/or of war between the Nations; on the other hand the new scholars were
a long time private people especially in the domain of philosophy but also
within purely scientific fields, - Descartes producing Analytical Geometry,
Pascal inventing the Calculus of probability and the mechanical computer
(cf. his Letter to the Swedish Queen accompanying the specimen he pre-
sented her with, comparing nobility by birth with nobility of the mind).
However, such scholars had to earn a living and many of them would seek
the sponsorship of princes (like Leonardo or Galilei), or an appointment
at one of the Universities (like Newton). Universities themselves shifted their
position from being more or less under the control of the Church to being
under the control of the State. But Universities were, and are still in prin-
ciple, endowed with a spiritual autonomy which Napoleon tried to destroy,
though he did not quite succeed. The French King Francois I in the 16th
178 Andre Mercier

Century had had the grand idea to found the "Royal College" (in our days
the "College de France") as the "second" Institute of Advanced Studies
in the world, the first having been the Museion created in Alexandria by
Ptolemy Soter at the beginning of the Hellenistic period Ord Century B.c.).
Science has always attempted at remaining a supranational business
by the establishment of a kind of unregistered world-wide corporation of
scholars structured according to the special domains. Neither the Church
and its prelates, nor the States and their Governments have ever liked it,
but they had to live with it, and when after two World wars Science, espe-
cially Physics, became such powerful components of the practical life
of nations in the fields of technology, hygiene etc., Science became a kind
of new power not to be neglected and requiring a statutary basis on the
national and international levels. Money had to flow into scientific enter-
prises in quantities comparable with what is needed for the maintenance
of armies.
This could be done along two lines. Either the officers of the State
would take over the organization of Science according to instructions
given by the Governments, or the scientists themselves would take the
initiative and elaborate the rules of their professional interrelations.
The money and the control over its sources however were not in the
hands of scientists; they were, - and are still, - in two hands: that of the
State, and that of industrial and other Companies. So these had to be
persuaded by the scientists to spend it on reasonable grounds if the latter
were to keep the initiative of the organization of research. Now, since man
is both a selfish and a generous animal, it could be expected that companies
and States would give "under conditions". In order to achieve a compro-
mise that would be satisfactory and favourable to the activity in science,
the speakers of the Corporation had to be endowed with at least three
qualities: to have achieved a high scientific standard, a high moral standard,
and to be skilled people in matters of administration. So, big men who had
built up or run large scientific institutions and who had proven to be in-
corruptible had to devote a good deal of their time, nay, of their career to
international understanding and to the management of scientific business,
some being more of the inspirators, others being more of the executors of
its organization, - sometimes being both at one time.
Organization has been found to be good when it is led under two
principles at a time: authority and democracy. They are not contradictory
when understood rightly. Among inspirators, Albert Einstein had a great
sense of democracy, because of a natural propension and because he had
grown up at College and University level in Switzerland, i. e. in the most
democratically minded country of the world. Among such as have been both
inspirators and executors, we might think of Arthur Compton, who spent the
last years of his life thinking and writing about the right ethics to suit scien-
tific activity. A third man whom I want to cite is Niels Bohr, from whom I
Birth and Role of the GRG-Organization 179

have learnt so much of the three qualities required. He and I entertained

important preliminary discussions before the official meetings took place
towards the creation of CERN.
When in the early fifties, the question arose in my mind what to do to-
wards the promotion of the physics of gravitation and the theory of relati-
vity, I attempted to win the moral and scientific counsel of three men: Ein-
stein (then at Princeton), Bohr (at whose Institute I had worked for two
years before Second world war) and Pauli (who was my elder colleague at
another Swiss Institution of higher learning). Einstein did not want to be
practically involved in the matter, his age and unsteady health and possibly
also his modesty preventing him to do anything. Bohr was rather reluctant
to join any team which would work out something, because - that was my
understanding - he felt that his connections were not with General Relativi-
ty and Gravitation but with Quantum physics and the concerns of nuclear
energy (cf. his Open Letter to the United Nations). Pauli on the contrary was
very happy about the idea to do something and was - against every expecta-
tion (his sarcasms against most undertakings of that kind were proverbial) -
most helpful and inspiring. We decided to have a "Golden Jubilee Confe-
rence" in 1955 in Berne where Einstein had around 1905 spent (in his own
words) "the happiest years of his life", and we agreed to consult one or two
scientists in the field who would be of my generation. So I contacted my
French colleagues with whom I had in former years attended especially
the advanced courses in Paris given by Elie Cartan: Andre Lichnerowicz
and the present Mrs. M.-A. Tonnelat, who, as we suspected, also planned
to do something. After some more consultation throughout the world,
the plans for the Golden Jubilee Conference were settled and, as those
who attended it will confirm, it was one of the most successful Conferences
which have ever been organized in the field of theoretical physics: Max von
Laue, who represented the Berlin Academy, even said so publicly at his
closing speech.
There are specific reasons to relate these things here, not of pride
but of pertinence to the whole matter of the previous and later development
of the physics of gravitation. Because, for those who have begun their
scientific career in the twenties or in the thirties, the recollection is still
tenacious that General Relativity was considered by most physicists bet-
ween the two World Wars as an academic affair to be left practically to
interested mathematicians. But having done a lot of thinking during the
years of the 2nd World War, especially on the applicability of new power-
ful methods of analytical dynamics, on the question of quantizability,
and on the epistemology of space-time and of the idea of unification in
physics, I had come to the conclusion that the time was ripe for a pretty
immediate expansion, even explosion of scientific work in the field that
would make out of General Relativity physics a main concern of the coming
generation of theoretical physicists. Pauli agreed with my arguments. So he
180 Andre Mercier

and I planned according to a very precise scheme a conference which would

concentrate on those fields which we thought were important. We did not
foresee, I must confess, precisely the rise of researches based on the disco-
veries connected with the singularities, pulsars, or black holes. But that is
not so important. What did matter was the recognition, that the field was
going to become one of the main fields of physical research.
Probably the French had had a similar presumption.
In any case, the importance of such work became so-to-speak overnight
internationally evident. Two years after the Berne Conference, Bryce DeWitt
had a smaller Conference in the United States which in a way was a reflec-
tion in America of what had taken place in Europe, while the French pre-
pared another bigger gathering that would make the whole undertaking to a
tradition and to the affirmation of an established organization.
Indeed, at the Conference of Royaumont in 1959, a final decision was
made to establish an International Committee with a Secretariate to support
all endeavour towards the promotion of the field throughout the world;
Academies and Science Councils of several countries had already supported
the efforts and IUPAP manifested great interest and became most helpful.
The Secretariat was established at the Department of Theoretical Physics
of the University of Berne with a few tasks written down on a kind of by-
laws; Professors Lichnerowicz and Tonnelat were elected as co-presidents
to chair the Committee until the next International Conference, a period
of three years was found reasonable to elapse between successive conferen-
ces, and the sigla "GRG" its symbol, meaning in either English or French
"General Relativity and Gravitation" or "Gravitation et Relativite Generale".
At the beginning, and during a number of years, the Committee remain-
ed a self-appointed group of people who considered one another as fairly
representative of the research going on in the field and of its world wide
distribution. IUP AP never contended it, and the praxis showed that it was
a good approximation. But an approximation of what? This is not easy to
say. Because contrarily to, say, Academies like the Royal Society etc. in
the 17th Century, our Committee had not received the approval of a "King",
nor had it received any donation which would make it not only spiritually,
but also financially powerful. Work at the Secretariat had to be done by
local staff and means, expenses of meetings of the Committee had to be
covered by all sorts of combinations. Small intermediary symposia held at
various places helped, e.g. the commemoration of Galilei's in Florence or
that of the General Theory in Berlin (to mention only two). To pay for the
publication of a Bulletin (or News-Letter), we had to raise small subscrip-
tions.
Royaumont was followed by Warsaw in 1962. At that Conference,
considered the third International Conference, Leopold Infeld succeeded
in the presidency of the Committee. Since then, an International Confe-
rence has been held every three years and called "GRn", Warsaw being
Birth and Role of the GRG-Organization 181

GR3. It has never been decided which Conference has been GR1: the Jubi-
lee Conference in Berne 1955 or the one in Chapel Hill 1957. But Royau-
mont was certainly GR2. This reminds of the Paris University, which was
founded 1150, recognized by the King in 1200, and received its final confir-
mation from the Pope in the year 1215.
Who says that the International Committee on GRG was finally recog-
nized at Warsaw? The following story will show that nothing like that was
actually the case! Anyhow, Infeld - inspite of his very poor state of health
- took very seriously the problems of the consolidation of our organization.
In particular, he maintained the opinion that owing to the tremendous
importance of the concentration of work and seeing the spreading of publi-
cations in so many journals ranging from pure mathematics over astronomy
to remote fields of physics, a new Journal had to be founded devoted ex-
clusively to GRG. He wanted to edit it himself, but he was ill and he also
hesitated on the ground, that Warsaw was outside the geographical area
where the modern "Latin" of science - even its modern "Greek" (every-
body will guess what is meant thereby) - are spoken. Before he died, he
gave me orally a sort of commission, not to drop the idea of a Journal
inspite of the reluctance shown by several members of the Committee. I
have considered it a duty to follow up the matter, and the first issue of
the GRG-Journal finally came out of Plenum Press in 1970, its preparation
having required a lot of work and securing the collaboration of a number
of referees.
But in the mean time, other events had occurred. One of them was,
that GRG, from being quasi-purely theoretical, had become an observa-
tional (not to say experimental) science with a very vast horizon. That
was a sensation for other physicists - not for relativists though. Astro-
physics - a chapter of astronomy - became "Relativistic astrophysics" -
practically a chapter of physics. It needed so its own special meetings. The
people in Texas had plans to establish research at the advanced level. They
called for a Symposium, which took place at Dallas in January 1963. The
procedure was repeated and became in the field of Relativistic Astrophysics
a series parallel to the Rochester Conferences in the field of Particle physics.
The series has since been called Texas Conferences, even though the confe-
rences may take place far away from Texas (like the "Rochester" e.g. took
place in USSR). It is an excellent supplement to the GRG-Conferences, for
if the latter cover systematically the whole field of GRG and so find their
suitable place among the big conferences in the sense defined by !UP AP
regulatives, the former allow to concentrate work on a most important
application at the boarder of physics proper and astronomy. Therefore,
those responsible for GRG have always welcome the Texas Conferences
and reciprocally. Especially, I have in former years attended these and
written big reports on them; but overwork and other duties have prevented
me to do so more recently.
182 Andre Mercier

Another very important happening, which of course involved not only

the activity in our field but the whole academic world of the time, was the
contention by young people that the academic "Establishment" of the old
was unbearable. We know now after ten years time that many of the claims
were justified, even though the Establishment was not that rotten. At the
same time, a growing tension between ideologies was felt on the political
plane. All that together culminated around the year 1968, but it had been
in the air already for several years. Both happenings came very suddenly
for most people, just as for the man in the street the coming into orbit
of the first artificial satellite by the Russians was a big surprise, though
the world of physics knew that the Russians were as able as any other
group to make it, if they had the motivation.
After GR3 in Warsaw, we had GR4 in 1965 in London, where Her-
mann Bondi was elected President. There, a light feeling of dissatisfac-
tion began to manifest itself. There were rumors of: "Why the Committee?",
"Who made it?" But nothing earnest happened, whereas the first serious
questioning was about the officiality of the relations between the GRG-
Committee and IUPAP. The suggestion that the Committee should become
a regular Commission of IUPAP was strongly opposed by the Committee
itself on the ground - which I think was correct - that the efforts of the
Committee were not only to promote GRG within physics proper but to
promote the collaboration in the field from Mathematics, Astronomy,
Physics, and even Philosophy, therefore IUPAP would not be a sufficient
frame for it. The result of that was a strengthening of the autonomy of the
whole organization and a clear and broad delimitation of its scopes, and the
Secretary General of IUP AP recognized fully the reasonableness of our
argument, without threatening in any manner to diminish the sponsorship
which had been granted so far.
Then came that period leading to 1968. It included even the armed
clash which happened in the near Orient and the armed occupation in Euro-
pe which divided opinions between East and West. I was the Rector of my
university at that time and had great responsibility, considering the students'
demands. Rectors and vice-chancellors throughout the world held close
together. The affairs at stake were considerable and their handling difficult.
Academic life and political ideologies were entangled, there were left and
right, young and old, tabula rasa and tradition, break and continuity; the
method used by many to contend everything had to be met with stubborn
serenity. The Committee was shaken by all that, partly from political and/or
ideological dissensions, partly from a growing rumor attacking it for not
being democratic, for not consulting the whole body of relativists in the
world and even for not having been constituted lawfully. It might for that
have broken down, but Bondi's firm yet understanding leadership kept
things and men together, proving that science is truly supranational and is
Birth and Role of the GRG-Organization 183

not to be distorted by non-scientific considerations, whereas the moral

foundations are proved by the promotion of the good and right. It was
both a problem of politics and policy, how to combine science and the
right.
GR5, which took place in 1968 in Tbilisi, stood under the shadow of
these events. But the organization recovered pretty quickly under the chair-
manship of the new elected President V. A. F ock, whom everybody respect-
ed. Yet the problem remained of the quarrel about the status of the Commit-
tee. The three years that followed required re-thinking our constitution, for
if the claim that the Committee had scarce foundation was justified, this
should be remedied, and if the other claim that its rejuvenation was due,
how should one proceed to realize it?
This led to GR6, which took place in 1971 in Copenhagen. The Con-
ference was a great success. Everybody was there, and a feeling of scientific
togetherness reigned. Everybody knew that Christian MQ\ller was going to
take over the chairmanship, that a democratization was planned, that the
Committee would undergo a change. Still, a storm burst out; a "general
assembly" had been convened to consider the constitution of an Interna-
tional Society for GRG, which would become the support of the old Com-
mittee as its own Governing body. One liked the idea, but the agitation
at the assembly was tremendous. Fortunately, an unperturbed Bondi could
announce that we had a new president, a secretary, a journal, and the prin-
ciples of a draft of a constitution. It ended in a general feeling of satisfac-
tion after applying these principles to elect new members to replace those
who retired on their own will or by lot.
During the years between GR6 and GR7, we were mainly busy fina-
lizing the constitution of the new society, and at GR7 in 1974 in Tel Aviv,
the Secretary was able to announce that a good many relativists had joined
as members. This was a great satisfaction to him, for it meant two things:
First, the legal situation had become unshakable, second the financial basis
of the whole GRG organization was insured.
The old International Committee on GRG was not altogether sup-
pressed, much rather, it remained as an autonomous international body
which has since been recognized officially by IUP AP as a so-called Interna-
tional Commission affiliated to the International Union, which gives it a
special status; IUP AP has the right to delegate a few representatives in the
Committee (at present there are four). But at the same time, the same
Committee became the governing body of the International Society, whose
General Assembly elects 24 members, as well as a president (the retiring
president being automatically deputy president) and a secretary. The Editor
of the Journal had changed in the mean time, for it had proved to be such a
job, that in view of other duties I had to pass it over to a colleague. At
184 Andre Mercier

GR 7, Nathan Rosen was elected President of the new Society and of the
Committee as such. The organization of scientific activity in the field of
GRG on the international plane has since been considered as secured and
indeed, it does work smoothly and efficaciously.
Twenty years of endeavours seemed a reasonable period and I decided
to resign as a secretary to be replaced at GR8 in Waterloo (Ontario), where
the General Assembly elected Peter Bergmann as the new president (another
former pupil of Albert Einstein's like the retiring president) and Alan Held
as the new secretary, who had already taken over the editorship of the
Journal.
All these informations may sound very administrative. But they are not
unimportant with respect to the good functioning of scientific research and
activity at large. 1979 will be the Centenary of Albert Einstein's birth. It is
something like a mile-stone at which we, scientists of the field, should be
able to take a rest and look both backwards onto the terrain which has been
travelled through so far and forwards onto the landscape which lies ahead.
With regards to achievements, specialists will be able to write beautiful
reviews on the various domains that have by and by crystallized during
more than seven decades, mostly after the Second world war. But a cente-
nary like that does not only mark a scientific advance, it also reminds of
human achievements which go back to the personality of the man under
consideration. These achievements are of scientific, moral, aesthetic and
even of an ineffable nature. A comprehensive philosophy radiates from the
whole situation in which we are, and since circumstances have made out
of me a man deeply involved in philosophy as a writer, a teacher and an
international officer, it seems appropriate to give the reader a hint as to
such items, more particularly in consideration of the experience described
so far.
Many scientists, who have not been able to study philosophy proper-
ly, have a tendency to beEeve, that science masters philosophy. Nothing
is more untrue. Philosophy is not a subject for science, it is not a science
either, only it has to be aware of scientific achievements (as well as of
other achievements of a nature different from that of scientific activities).
Philosophy is not either, as some believe it to be, a substitute for a
religion that would have become obsolete. Man is a peculiar being, because
he is deeply affected by the fact that he realizes that he does not know
much, especially not why he is down there on the middle planet where he
encounters such a mess and so many hindrances on his path. So he is very
much distressed. But at the same time, the regular relative motions of the
celestial bodies, the symmetries in crystals and plants, and many other
features of things attract his attention as being so very beautiful, good,
true and loveable, that he is filled with a joy at the prospect of being able
to understand them and of coming nearer some sort of perfection him-
self.
Birth and Role of the GRG-Organization 185

There can be no doubt that Albert Einstein has been one of those who
felt most deeply the impress of that dichotomy. Now there are two and only
two possible ways to walk from the original situation of distress towards
that state of perfection which every one would like to reach. One consists in
never being afraid of the stones and the dirt which injures and stains our
bodies, for there is a voice that tells us that this is of no moment. The
other one consists in examining at each step the situation arrived at in
order to clear it of all the obstacles which seem to prevent an understanding
of the next possible step. Both ways are there to prove our existence and
to ground our freedom.
Everybody knows that Descartes exclaimed: I think, therefore I am.
Well, to think is to cautiously walk along the second path; a thinker is a
philosopher, and philosophy is the second possible way to let our existence
show itself from our original si~uation of distressing ignorance towards one
of growing comprehension. But it is less well known that Descartes had an
experience which engaged him to also try the first possible way. Of course,
the cogito, ergo sum does not say a word of what is to be done at each step
of the cogitation. This is up to each philosopher to decide. Descartes worked
very hard elaborating his career, inventing e.g. Analytical geometry which is
recognized by all mathematicians, but proposing a physics that has been
found totally inadequate.
A few years younger, Pascal never explicitly made a comparable claim.
But all his life may be said to have illustrated one other claim which I might
formulate like this: I (i. e. Pascal) believe, therefore I am. That is what leads
one along the first path, religion if you please, i.e. the faith in God. But
such a credo, ergo sum does not explicate any of the injuries which one
experiences by being a religious person. Pascal contributed very valuable
advances in mathematics, physics and theology, and was a very great poet
even though he wrote in prose, for his French is of the most beautiful kind
that has ever been written, and he was a great philosopher too.
Descartes had many followers: Spinoza, Leibniz and others. Pascal
was more of an example and a source of inspiration, and Leibniz respected
him highly; but he had no followers.
I have not chosen my instances from Antiquity, but from the so-called
Modern times, for a very specific reason. One knows since the work of
Arrhenius on the history of science that the rise of modern science, espe-
cially Newtonian physics, would not have been possible without the Judaico-
Christian conception of Time having replaced the antique concept. For Plato
had a cyclic idea of time (which had possibly come from ancient wisdom in
India through Pythagoras), whereas the idea of historical linearity came to the
awareness of thinkers with the Bible, including that of a creation, i. e. a zero
time and, with Christian eschatology, even an end of time. Science has never
said a word about a possible end, but it has been built upon the new concep-
tion of time as of the independent variable imposed on everything and upon
186 Andre Mercier

which everything else, without any exception, depends. This is even so in

Relativity physics, for- contrarily to what is said in many text-books -
space-time is not a (generalized) space, it is (generalized) time.
Einstein makes no exception. If I were to compare him with someone, I
should say that he is in a line with Descartes. But Descartes - not knowing
much in his days - made the mistake to try to elaborate a physics in space
by filling it with vortices, not realizing that space is not a physical container,
but a mathematical construct; Einstein - who knew a lot more owing to
three centuries of good physics - elaborated a physics in "time" (in space-
time), and it worked.
But Einstein is not only Cartesian in physics, he is so too philosophi-
cally, even more so: he is Spinozean, with regards to matters of morals and
of belief. I have neither talked nor had correspondence with him on that,
but it is visible from his life and works, and indeed generally recognized.
He even felt for something which does not seem to have retained much of
the attention of the Cartesian proper (except Leibniz): he was deeply artisti-
cally minded too, in a manner quite parallel to his scientific, moral and even
his mystic minded-ness, which may be said to be exhibited in the music of
J. S. Bach. Let us not forget that Bach's music is also a reflection of Christian
mystique. Since I have named before the other great physicist of our Centu-
ry, let me add that Niels Bohr should for his part rather be compared to
Pascal, as I know from personal conversations with him on these matters
(and as has been explained in a publication by Oscar Klein), than to Descar-
tes.
Men of that size, who have reached a summit within the limits approa-
chable in their own time and their particular environment in the intellectual
sphere, in moral respect, in relation to art and in the contemplation of the
divine, exceed our minds and pledge us to modesty.
Hence, if anything is being done, elaborated, organized and decided
so-tospeak "under their sponsorship" or "in their name", it should be done
by keeping these things in mind. Einstein knew for instance, even though
perhaps he did not explicitly say so, that technology has both a scientific
and a moral aspect, that neither the morals are to be judged upon by scien-
tific criteria nor the sciences to be measured by moral standards, but that
morals and sciences have to join into an equilibrium in a harmonious en-
counter which would constitute a fully responsible technology. He knew,
generally speaking, that non-scientific matters are to be neither confirmed
nor rejected on the ground of scientific reasoning, for science is not habili-
tated to do so since its scope is limited by the nature of its approach of the
real.
There is a German word which has not an easy translation into English
or French, viz. the word "Sachlichkeit". It should not be translated with
objectivity; for, objective is only the scientific procedure, where the relation
between the scholar and the object of his cognition goes one way and
Birth and Role of the GRG-Grganization 187

assumes the greatest possible independence towards his object. Objectivity

is falsely defined by intersubjectivity, for in art or in morals there is as much
intersubjectivity as in science. I have repeatedly translated in my work the
word "Sachlichkeit" with pertinence. Its negation: impertinence means,
that one mixes with things that are not one's business. Precisely, it is not
the business of science (it is not pertinent) to mix with non-scientific mat-
ters. Any scientific analysis of a work of art for example destroys it comple-
tely as a work of art. Only at a higher level do the true and the beautiful
come into an interplay which is the essence of a kind of architecture.'
The four fundamental, irreducible values - philosophers call them
cardinal - are the true, the beautiful, the good and the sublime. Their
promotion can be shown to be the reward of a combined theoretical and
empirical enterprise. When we buy a beautiful picture, we pay for it, which
is a quantity, but that quantity is a measure of the recognized quality.
Quality and quantity are not opposed as contradictory, they reflect each
other; there is no quality that is not finally measured, and there is no quan-
tity that is not the measure of some quality. Of course quantity should not
be confused with abstract number or quality with extrinsic appearance.
Value is the coincidence of quantity and quality.
In all contributions to the promotion of international understanding
and working in science and, more generally, in philosophy, I have been
led by such considerations as have just been laid down. International organi-
zation in science, particular elements of which are Conferences, is not just
an activity to be judged by its mere scientific achievements. It is a technic
of its own, and a technic is always an encounter like one between science
and morals. When in technology, for instance, an engineer builds aqueducts
in order to bring pure water through the land into people's homes in the
towns, he must know statics and hydraulics as scientific disciplines, but
he must also understand the needs of peasants and their crops from the
land and the necessities of enlarging the hygienic possibilities of people
and the like, which are all of moral nature. When organizing collaboration
in science on the international level, e.g. by inviting to conferences, it is
not enough to hire managers in order to insure its success; one needs highly
skilled specialists in every respect. It does not work e.g. just to declare a
meeting open to all; openness is a moral criterium of democracy of a quan-
titative nature, but its misuse happens again and again; the corresponding
qualitative nature is to be achieved on very specific grounds. Many a parti-
cipant at international conferences is not aware of the delicacy which the
realization of such harmony requires. Policies have to be discovered, learnt
and applied.

1 I shall not quote books and articles where these things are explained at length.
188 Andre Mercier

The year 1979 will not pass away without there having been a number
of Einstein Centenary Celebrations throughout the world. Here again, pride
and prejudice might distort the accurate picture. Not only those who under-
take them, but also those who will profit by them should refrain from falling
into such a mistake.
History has to be written impartially.
 189

Reminiscences ofAlbert Einstein from 1908 to 1930

Walther Gerlach

Reminiscences from personal experience of a man who is outstanding

in his field are inseparable from the history of his science and from the
person's own development.
The first contact with Einstein, and probably the most decisive for my
future life, was of an impersonal nature. It was the end of April, 1908. I
wanted to study philosophy and mathematics in Tiibingen and asked the
philosopher Erich Adickes for his advice. "Mathematics is fine, but you
should study physics, too. If Kant were alive today, he would base his ideas
on Einstein, not Newton." I had never heard the name before, and had no
idea about physics, while I had enjoyed particularly good instruction in
mathematics under Professor Biicheler at the Humanistische Gymnasium in
Wiesbaden. So I attended the lecture and practicum of the physicist Frie-
drich Paschen and was so fascinated that I spontaneously gave up the idea
of studying philosophy. Not until later did I understand the amazing advice
given me in 1908 by a philosopher. The mathematician Alexander Brill and
Privatdozent for Physics Richard Gans had certainly discussed the - as we
call it today special theory of relativity, but at the time only two years old -
in their "Tuesday discussion group". Apparently this had somehow impres-
sed Adickes, who was concerned with Kant's scientific thoughts.
The study of physics consisted of a two-semester lecture in "Experi-
mental Physics", five hours a week, and a four-hour practicum twice weekly,
with a semester lasting from the end of April till the beginning of August and
from the end of October till the beginning of March. Beginning with the
third semester one enrolled in three lecture courses in mathematical physics
(more mathematics than physics) and one-hour special lectures held by Pri-
vatdozenten Richard Gans and Hans Happel. I heard about the theory of
relativity for the first time in a lecture given by Gans in connection with the
increase in mass with velocity of cathode-ray particles. In my final oral exa-
mination on February 29, 1912, Brill wanted to know more about the
theory of relativity from me than I really knew. One acquired training in
physics essentially from Paschen's practicum with the help of his assistants
Gans and Paul Gmelin; whoever really wanted to devote himself to physics
enrolled in the practicum for several semesters. He got to use better appara-
tus, and was handed special-prints and monography to be studied. In this
way one was led to premature independent study of original texts.
190 Walther Gerlach

The state of physics at that time can be described thus: there were
generally interesting special areas, such as longwave infrared, gaseous disch-
arges, spectroscopy, radioactivity, canal rays, which were being worked on in
various institutes. The theoretical foundations were thermodynamics, the
kinetic theory of gases, electromagnetism, and the electron theory of the
electrical and optical characteristics of matter. But there were no basic
questions and certainly none concerning relativity and quantum physics.
Through the papers appearing for Planck's lecture on the "Theory of Thermal
Radiation" I became acquainted with Einstein's light-quantum hypothesis 1 .
This became the object of many fantasies with Ernst Back, who having
finished his law studies changed over to physics, and who became well-
known for the Paschen-Back-effect. I do not recall Paschen's opinion on the
matter. To him, as a spectroscopist, the problem of the atom lay in the
"monochromatic undamped Planck oscillator", as he so often stated. After
all he motivated Paul Gmelin and me around the year 1912 to investigate
the pressure of light, searching for the deflection of a jet of iodine vapor
by intense transverse radiation. Otto Robert Frisch 2 first proved this "Ein-
steinian light-quantum momentum" in 1933 in absorption and emission.
After the publication of the experiment of James Franck and Gustav
Hertz 3, which was later to become famous, Back and I went to Paschen
quite upset: it was the inverse photoelectric effect, namely the release of a
light-quantum by means of an electron of the same energy. Paschen rejected
this brusquely, saying that the connection between the emission of spectral
lines and ionization had not been determined yet. We tried to link up every-
thing with Bohr's theory of the atom 4, which was not given so much atten-
tion elsewhere (in Berlin, for example), since Paschen had recommended it
to us in the summer of 1913 with the prophetic words, "That is physics for
the next thirty years." He had been impressed by the first interpretation
of the Runge-Paschen-Ritz term and term-difference law for spectral lines.
He never mentioned it again until Sommerfeld's theory of the fine-structure
of spectral rays and in his own measurements, which far exceeded the limi-
tations of accuracy for that time, seamed to settle the matter. When he
wrote the famous Annalenarbeit with the curious title "Bohr's Helium Lines"
in 1916 (Sommerfeld had come to Tiibingen for several days) he repeatedly
added to his admiration for the theory by saying "Now for the first time I
have experimentally proven the theory of relativity." Before this I had never
heard a word mentioned about it at the institute.

A. Einstein, Ober einen die Erzeugung und Verwandlung des Lichts betreffenden
heuristischen Gesichtspunkt; Ann. d. Phys. 17, 132, 1905.
2 O. R. Frisch, Experimenteller Nachweis des Einsteinschen StrahlungsriickstoGes,
ZS f. Phys. 86, 42,1933.
]. Franck u. G. Hertz, Ober die Erregung der Quecksilberresonanzlinien durch Elek-
tronenstiiGe, Verh. d. D. Phys. Ges. 16, 512,1914.
4 N. Bohr, On the Constitution of Atoms and Molecules, Phil. Mag. 26,1,1913.
Reminiscences of the Albert Einstein from 1908 to 1930 191

The discovery of X-ray interferences and the proof of the atomic crystal
lattice, the vivid idea and theory of Max Laue s and the lucid experiment of
Walther Friedrich and Paul Knipping in the summer of 1912, excited the
physicists more than the other discoveries of that time, such as the electro-
optical Stark-effect, the atomic nucleus by Ernst Rutherford or the supra-
conductivity by Heike Kamerlingh Onnes. It enlivened the discussion about
Einstein's light quantum theory: This had previously been brought into con-
nection with the concept of X-ray corpuscles or shock waves which released
and excited high energy electrons. In agreement with Paschen and Meyer I
experimented with interferences of Alpha-corpuscles transversing crystals.
In a vacuum tube we arranged within a few centimeters a Bragg-diaphragm,
which made Alpha-rays parallel. The rays fell through a mica foil onto a
fluorescent screen. But it was not possible to find symmetrically ordered
scintillations outside of the ray. At this point Edgar Meyer suggested testing
a fundamental question of Einstein's theory of the photoeffect: whether the
release of an electron occurs spontaneously with the absorption of light or
whether it occurs after an accumulation time during which radiation energy
is absorbed 6. Ultramicroscopic metal particles were bombarded with sharply
focused ultraviolet rays in an Ehrenhaft-Millikan-condensor 7 . Delays ranging
from fractions of a second to many seconds were recorded from the onset of
bombardment to the first observed changes in charge. Upon repeating the
experiment these delays varied around a mean values, which in turn depended
on various experimental conditions. The alterations in charge amounted to
one or several electronic charges. H. Joffe carried out similar experiments of
his own at Rontgen's Institute. We reported on our results and their possible
significance at the spring conference of the Society of Physics held in Zurich
in 1913. At this occasion I became acquainted with Einstein. I had already
met some of the physicists whom we revered: Heinrich Rubens, Carl Runge,
Johannes Stark, Pierre WeiB, Willy Wien. The 34-year-old Einstein was of
quite a different nature. I know very well that the recollection of earlier
sentiments is a very touchy subject; autobiographies give us enough examples
of this. The recollection of an event is a different matter; an event which
revealed quite unexpectedly and surprisingly a new human and scientific
insight, which was reinforced year after year in my sporadic meetings with
Einstein.

W. Friedrich, P. Knipping u. M. Laue, Interferenzerscheinungen bei Rontgenstrahlen,

Sitzungsberichte der Bayer. Akad. d. Wiss. 1912, Seite 303. Theorie: M. Laue, eben-
da,S.363.
6 P. Debye u. A. Sommerfeld. Theorie des lichtelektrischen Effekts yom Standpunkt
des Wirkungsquantums. Ann d. Phys. 41, 873,1913.
7 E. Meyer u. W. Gerlach. Uber den photoelektrischen Effekt an ultramikroskopi-
schen Metallteilchen. Arch. d. Geneve (4) 35, 398, 1913. Ann. d. Phys. 45, 177,
1914. Ann. d. Phys. 47,227,1915.
192 Walther Gerlach

It was the first scientific conference I had attended where I was also to
present a lecture and I was anxious and nervous. Einstein was pointed out
to me: inconspicuous, rather casual in posture and attire, he sat in an audi-
torium (small by modern standards) in which everyone knew everyone else.
Occasionally he interrupted a lecture with a pointed question, or debated
a matter with a subdued voice marked by a slight south German accent,
always clearly worded even if improvised. In the course of the conference
we met in smaller circles at lunch or during breaks, once at the house of a
patron, where Einstein played the violin. I noticed at once his extreme
modesty. As the saving went in Zurich in those days, "Be a man, eat
Schiibli and smoke Stumpen", and such a man he was. "I go fourth class,
and still get where I'm going." He spoke so naturally that one forgot his own
inhibitions. He continually made sure that he had understood your point
of view by asking questions. He seemed to be interested in everything, he
"thought aloud". If he rejected an idea, he did it kindly, making it milder
with a witty comment. First and foremost he was concerned with the matter
at hand, free from prejudice. Arrogance and complacency were foreign to
him - he saw in them the root ot the nationalism which he hated. He spoke
about this frankly, as I have only heard since in the fall of 1914 from my
teacher Friedrich Paschen.
I remember details of a longer conversation about our experiments
together with Edgar Meyer. The much debated "sub-electrons" of the
Viennese physicist Felix Ehrenhaft and his ideas on photophoresis in con-
nection with the light-quantum issue came up for discussion. Einstein asked
for details as to why we rejected the sub-electrons 8 . Our experiments showed
unequivocally the atomic structure of electronic charge; however, the
absolute values varied from particle to particle on a wide scale. From this
we concluded the non-validity of the Stokes formula which we had used in
calculation. His opinion on our experiments concerning delay or accumula-
tion time of the photoelectric effect was that this must be a secondary effect
(as we had also concluded). An ionisation of the surrounding gases by means
of the ejected electrons impossible. It was a possibility that ejected elec-
trons attached themselves to neutral gas molecules and ions were formed
- which after having been transferred back to the particle within a short
time by means of Brownian motion or through electrical attraction -
prevented the observation of the release of an electron, and so an accumu-
lation time appeared to occur. Some time later we received a note from Ein-
stein (written on a slip of paper) with some formulas and the question
whether or not we had proved our hypothesis. So we measured the delay
time of the same particle while reducing the surrounding atmospheric pressure

8 E. Meyer u. W. Gerlach, Ober das Elementarquantum der Elektrizitat, Ann. d. Phys.

48,718,1915.
Reminiscences of the Albert Einstein from 1908 to 1930 193

to about one-tenth, and we found the expected drastic decrease of the

delay time. These observations served as a basis for testing Stokes' formula 9 .
During our conversation in Zurich Einstein suggested that I should try
to clarify the difficulty that still existed in his light quantum interpretation
of the photoelectric effect. He had to separate the work of a light quantum
into two parts - the kinetic energy of the escaping electron and an energy
depending only on the properties of the irradiated metal of the same order
of magnitude. Here the negativ criticism of Planck set in: to explain the
failure of the simple relation (light quantum equals electron energy) Einstein
had a second hypothesis at hand. Assording to him the difference of this
additional value of two metals was the same as their contact potential;
I was to measure this with the aid of the "Einstein multiplier"IO developed
in 1908. Paschen was so enthusiastic about it that he bought a not inexpen-
sive apparatus from the mechanician Habicht in Basel. In spite of all efforts
we did not succeed in obtaining reproducible results. Today we know the
reason why: at that time pure metal surfaces were not available.
The observation of Brownian motion in the aforementioned experiments
with ultramicroscopic particles and Einstein's decisive work 11 (as well as its
supplement by Smoluchowski which had just been published) induced me to
try to measure the oscillation of a torsion-balance (silvered mica sheet sus-
pended on a very thin quartz filament). The experiments were interrupted
when I was recalled to military duty and were taken up again in 1926, this
time with improved devices 12. As continued by Eugen Kappler 13, they
eventually led to the first absolute measurement of the Avogadro-Loschmidt
number.
I did not see Einstein again until after the First World War. As a physicist
for the Bayer paint factory in Elberfeld I took part in the Assembly of
Scientists and Physicians in Bad Nauheim, which was held in conjunction
with the yearly conference of physicists. A dispute between Einstein and
Philipp Lenard about the "general" theory of relativity was planned. I was
familiar with Lenard only through his publications, since we considered his

9 E. Meyer u. W. Gerlach, Ober die Giiltigkeit der Stokesschen Formel, Festschrift fUr
J. Elster und H. Geitel, Verlag Vieweg 1915, Seite 196.
10 A. Einstein, Eine neue elektrostatische Methode zur Messung kleiner Elektrizitats-
mengen, Phys. ZS. 9, 216, 1908.
11 A. Einstein, Ober die von der molekularkinetischen Theorie der Warme geforderte
Bewegung von in ruhenden Fliissigkeiten suspendierten Teilchen. Ann. d. Phys. 17,
549, 1905. -- Bez. der benutzten Methode s. M. von Smoluchowski, in Vortrage tiber
kinetische Theorie der Materie, Gottingen 1914.
12 W. Gerlach u. E. Lehrer, Ober die Messung der Brownschen Bewegung mit Hilfe einer
Drehwaage, Naturwiss. 15, S. 15,1927.
13 E. Kappler Versuche zur Messung der Avogardo-Loschmidtschen Zahl aus der
Brownschen Bewegung einer Drehwaage. Ann. d. Phys. 11, 233, 1931.
194 Walther Gerlach

early works on electron beams to be examplary models for experimentation.

I was all the more astonished that he participated in the recently provoked
repulsive agitation aimed at Einstein, particularly since the English expedi-
tion in 1919 to study the eclipse of the sun resulted in reliable - if only
qualitatively - evidence that stellar light is deflected in the vicinity of the
sun. The discussion between Einstein and Lenard was conducted by Max
Planck on September 23, 1920, in the large casino Bad Nauheim. Only a
small minority of the many present were physicists. Lenard continually
reffered to the undefined term "Anscbaulicbkeit" (clearness); Einstein put
critical questions. Planck tried to quickly suppress excited tangents leading
into personal criticisms. The audience was reserved as far as pro and contra
statements were concerned. I do not recall exactly the details of the various
arguments presented - they have been printed and quoted much too often
since then. But the impression with which I left the meeting remains un-
changed: It was not productive, even disgraceful for Lenard. The opponents
of Einstein who had been hoping for a tribunal had to be disappointed; the
sensation, which was occasionally reported on later, simply did not take
place. Perhaps it is significant that the whole matter is mentioned only
incidentally in the correspondence between Born and Einstein. Certainly
this discussion was no daily occurrence, but scientifically speaking it was
unproductive. Ideological and personal motives are totally out of place in
a scientific discussion concerning the possibilities and significance of a
totally new mode of thinking. It was too early to speak of proof or disproof
of the general theory of relativity by means of physical experimentation.
The opposition became more and more irrelevant. Only very few reputed
physicists implicated themselves by openly supporting Lenard - perhaps in
rightfully defending the work of Lenard and Johannes Stark, who had won
the Nobel Prize, or under the guise of scientific caution. But a flood of
articles and brochures criticizing the theory of relativity appeared, primarily
from philosophers who claimed to have a monopoly on "time and space",
in the sharp words of Einstein. As far as I can recall, Lenard raised no ob-
jections to the special theory of relativity in Bad Nauheim. Physicists such
as Peter Paul Ewald and Max Laue wasted time trying to clarify misunder-
standings. Ewald pointed to the term "relative" in a lecture: As Newton
had once done, so Einstein was now making absolute statements based on
physical facts.
Physics transcended such differences, but in my generation the shadow
remained which had been cast over the otherwise happy twenties, and which
had fostered a suspicion of all things philosophical.
In the fall of 1920 I accepted a teaching position for "Higher Experi-
mental Physics" in Richard Wachsmuth's institute in Frankfurt and was
recruited by Max Born and Otto Stern to help with their experiments with
atomic beams at the Theoretical Institute. I had gathered experience in
Tiibingen with Dunoyer optical atomic beams, and began now with magnetic
Reminiscences of the Albert Einstein from 1908 to 1930 195

ones. Stern 14 then made the suggestion of investigating the so-called direc-
tional quantization with an experiment which was basically very simple:
the deflection or splitting of a silver atomic beam in an inhomogeneous
magnetic field. For want of any financial aid at a time of beginning infla-
tion, Born supported us from fees he had received by lecturing on the theory
of relativity. Due to great difficulties it was often discussed if the attempt
was even worth the effort. Niels Bohr and Max Born expected proof of
extreme directional quantization, Arnold Sommerfeld at most a semi-
classical result, Peter Debye maintained that experimental evidence obtained
from magneto-mechanical experimentation was impossible, and Fritz Haber
gave us a large sum of money from the Hoshi- Foundation to continue experi-
mentation and not give up hope. Stern and I always used to say, "The autopsy
will tell", where Stern favoured a classical rather than a quantum interpreta-
tion. Born kept Einstein posted, and he provided us with money for a strong
magnet after the first promising results became known, and helped us with
his optimism. When the experiment finally succeeded 15 he wrote the follow-
ing lines to Born: 16 "The most interesting thing at present is the experiment
of Stern and Gerlach. The alignment of atoms without collisions cannot be
explained with our present understanding of radiation. An alignment should
rightfully take over a hundred years. I did some calculations with Ehrenfest.
Rubens considered the experimental result to be absolutely certain."
At this time Einstein suggested that I should concern myself with an
entirely different issue- the question as to whether matter in motion creates a
magnetic field (evidence of such experiments were to be found in Michael
Faraday's diary). Measurements alongside streams and waterfalls were meant
here. I was to give up my academic work for the time being. After much deli-
beration I could not make up my mind to do this.
During some such conversation Einstein mentioned that the flame of a
burning candle would disappear in free fall, since it was dependent on con-
vection and therefore on the acceleration of the earth. I answered, "There-
fore a gas-filled incandescent light bulb becomes brighter when falling." I
have shown these two experiments 17 regularly in lectures since my time
in Frankfurt. A flame contained in a large, closed bell jar retreats to a glow-
ing area near the wick when falling freely, while a slightly glowing filament
in the bell jar gets brighter.

14 O. Stern, Ein Weg zur experimentellen Priifung der Richtungsquantelung im Magnet-

feld, ZS f. Phys. 7, 249. 1921. Lit. iiber R. Qu. siehe A. Sommerfeld, Atombau und
Spektrallinien 1921.
15 W. Gerlach u. O. Stern, Der experirnentelle Nachweis der Richtungsquantelung im
Magnetfeld, ZS f. Phys. 9.349.1922, Ann. d. Phys. 74, 673,1924.
16 M. Born, Albert Einstein - Max Born Briefwechsel, S. 103 Nymphenburger Verlag
1960.
17 W. Gerlach. Einige Vorlesungsversuche. ZS f. Phys. u. chern. Unterricht 50, 139,
1937. Jahresbericht d. Phys. Verem Frankfurt/M., 1963, S. 63.
196 Walther Gerlach

During the twenties I met Einstein occasionally in Berlin. I can remem-

ber an episode from a colloquium in the Institute of Physics. He attracted
no particular attention. It was similar to the time in Zurich, but the audi-
torium was larger and he was sitting in the first row. Walter Nernst was
lecturing and Einstein interrupted with an objection that led to a rather
agitated discussion between the two of them. The chairman - perhaps it
was Heinrich Rubens - broke it off, requesting that the gentlemen continue
their discussion after the lecture had been concluded. Nernst continued
speaking while Einstein sank his head, winding a lock of hair as usual.
Suddenly he sat up, raised head and hand - he had very apparently grasped
the matter. Nernst noticed this, took a few steps towards Einstein, made a
deep bow, and then turned again to his blackboard.
Another time we were all enjoying a casual evening together. Einstein
appeared to be quite depressed: He had been informed by the Patent Office
that the principle which he and Szilard had used to invent a refrigerator was
already registered. - "Well, if they had told me that about the theory of
relativity, then ... - but ... "
In the spring of 1926 Einstein published his shrewd and provocative
treatise concerning the wave-particle-duality of matter, "Uber die Inter-
ferenzeigenschaften des durch Kanalstrahlen emittiehen Lichts. "To be able
to appreciate its fundamental importance one must keep in mind the views
on this basic question at that time. The validity of Planck's formula for black
body radiation could not be doubted despite recent objections by Nernst
following the measurements of Rubens and Michel. Einstein's prophesy of
1905, "that Mr. Planck had introduced a new hypothetical element into
physics", had been fulfilled. It explained the various energetic interactions
between electromagnetic radiation of all frequencies with matter in all states
and with free electrons: the emission and absorption of spectral lines (in
connection with Bohr's quantum condition and Sommerfeld's "Atomic
Structure of Spectral Lines"); the Compton-Debye effect (related to Ein-
stein's equivalence of mass and energy); and the experimental proof by
Walther Bothe and Hans Geiger of the validity of the conservation of energy
in each single process for the transformation of quantum energy to kinetic
energy (in opposition to the theory of Bohr, Cramers, and Slater). The
existence of the high-frequency limit of X-rays spectra (in connection with
the precision determination of Planck's constant h) and the importance of
energy quanta for the thermal behaviour of matter at low temperatures
could be understood. The first papers on quantum mechanics - Born, Jor-
dan, Heisenberg, Dirac - had just appeared, but not yet Schrodinger's wave
mechanics. But on the other hand, on the basis of various experimental
results, the propagation of light in the form of long wave-trains could no
longer be doubted; the existence of de Broglie's monochromatic matter
waves had not been proven yet.
Reminiscences of the Albert Einstein from 1908 to 1930 197

The dual nature of light had been considered already one hundred years
before. Thomas Young's transverse undulatory theory (further developed by
Fran~ois Arago and finally by Augustin Fresnel) stood in opposition to the
"valid" Newtonian concept of (quasi magnetic) polarized light corpuscles,
which was generally accepted and apparently guaranteed by Etienne Malus'
discovery of "polarization". Then Jean Baptiste Biot claimed the explana-
tion of interference using corpuscles as the most urgent task of physics.
Fraunhofer stated among others on the basis of the first absolute wavelength
measurements using his diffraction grating in air and water ~ "Interference
will always exist. Whoever can imagine anything other than a wave with
these characteristics, must adapt it to his view".
The "either-or" of that time stood in opposition to the "as well as" in
Bohr's principle of complementarity and proven experiments on wave-
particle duality. Experimenters said, when asking about energy relationships
the answer is in terms of "quanta"; when asking about a radiation field, the
answer is in terms of "waves". Einstein 18 posed the sceptical question to
the latter: Why doesn't light, which is emitted from canal rays (that is, from
rapidly moving atoms), create interference? However, he demonstrated that
"the failure of the classical undulatory theory was almost out of ques-
tion" because of an experiment derived "without special presumptions, in
particular also of those concerning the undulatory theory". Light from canal
ray atoms (of uniform velocity) creates and interference pattern of "rest" if
one of the mirrors is tilted at an angle depending on the velocity of the canal
radiation in the Michelson-interferometer. With "Einstein's tilting-mirror
experiment", radiation interferes which is emitted at varying times by par-
ticles moving in the same direction. Einstein sees the special significance of
this consideration fundamental for his reasoning, into which I had the
honor of having insight - in the fact that it "leads to a convenient predic-
tion of the expected interference patterns". After only three months Emil
Rupp 19 submitted the experimental results which "fully proved the theory",
as Einstein emphasized in the concluding statements. The "total validity of
the undulatory theory (according to Bohr and Heisenberg)" and the invalidity
of the presumption "that the radiation field which determines the inter-
ference can be produced in a momentary procedure, as proposed by the
quantum theory", were both fundamental for the progress of physics. Willy
Wien urged his student Harald Straub to verify the Rupp experiments. The
result corresponded to his reasoning, yet he was sceptical due to a previous
work of Rupp's on phosphors. After Wien's death Eduard Riichardt and I

18 A. Einstein, Dber die Interferenzeigenschaften des durch Kanalstrahlen emittierten

Lichtes. Sitzungsberichte der Preuss. Akad. d. Wiss. 1926, XXV, S. 334.
19 E. Rupp, Uber die lnterferenzeigenschaften des KanaIstrahIIichtes. Sitzungsberichte
derPreuss. Akad. d. Wiss. 1926, XXV, S. 341, und Ann. d. Phys. 79 1926, S. 1.
198 Walther Gerlach

held the responsibility for Straub's experiments. After tedious attempts on

the part of Straub with the use of all possibly attainable means we three
came to the conclusion that the experiment as conceived by Einstein could
not be carried out. 20 It was still not possible at that time - particularly in
the way described by Rupp - to produce canal rays of such homigeneous
velocity that the velocity would be compensated for by the specific angular
tilting of the mirror.
Rupp 21 thereupon published for the first time "photographs" of the
interference of moving light sources accompanied by detailed accounts of
mirror distances and focal lengths which corresponded to Einstein's test
conditions. In the meantime doubts arose as to whether the image condi-
tions presupposed in Einstein's theory and used by Rupp were necessary at
all. The result of Einstein's consideration seemed to be independent of these
conditions; yet they were essential for the testing of the theory. Besides,
we discovered that there was a mistake in the experimental sketch in Ein-
stein's paper of 1926: the direction of rotation of the mirror would have to
be just the opposite of that given, according to the text and in reality. Rupp
had used the wrong turbing direction in his experimental arrangement. We
notified Einstein and he wrote only briefly that he would be meeting me
shortly at the Solvay-congress in September, 1930 in Brussels and could
discuss it at that time.
We made an appointment for a certain day to have coffee together after
lunch. Shortly before lunch Einstein took me aside saying he had to go to
London immediately. Mr. So-and-so had summoned people together who
were to donate money for the Jews. If Einstein did not speak to them, there
would not be enough donations. After a few troubled remarks about the
anguish of the Jews throughout the world and his concern about a new
Zionistic nationalism, I begged him urgently to review the meaning of this
mistake in thought in this paper which was of such importance for him and
for us. Tormentedly he replied, "I can't do anything about it, but I must
help ... Give Riichardt my regards and please continue your experiments."
I remember so clearly every detail of that conversation which took place in
a curtained alcove: the disappointment that our major problem had not been
discussed, Einstein's worried face, his doubt, and the cordiality of his fare-
well, which was to be the last.
So I returned to Munich having achieved nothing. But we could not.
make up our minds to publish our criticisms of Einstein's paper, even after
Rupp brought out two pictures a short time later as experimentum crucis

20 H. Straub, Uber die Koharenzlilnge des von Kanalstrahlen emittierten Lichtes. Ann.
d.Phys. 5, 1930,S. 644.
21 E. Rupp Erwiderung zu der Dissertation von H. Straub "Uber die Kohilrenzlilnge des
von Kanalstrahlen emittierten Lichtes". Ann. d. Phys. 7,1930, S. 381. Antwort auf
die Bemerkung von H. Straub, Ann. d. Phys. 8, 1931, S. 293.
Reminiscences of the Albert Einstein from 1908 to 1930 199

for his proof of the mirror experiment: interference with Eintein's image
conditions, no interference without them. The wrong tilting direction in
Einstein's and Rupp's drawing did not seem to us to be sufficient proof of
fraud. While experiments to produce canal rays of homogeneous speed in
our institute continued, Riichardt succeeded in deriving the Einstein-experi-
ment within the framework of another project. The derivation precluded
once and for all the image conditions. In 1935 22 we published our correc-
tions of Einstein's derivation and drawing with the statement that Rupp had
not conducted the tilting-mirror experiments and that the interference pat-
terns had been falsified. Some time later C. Ramsauer made the same claim
in regard to other works of Rupp.
In 1937 - 38 Heinz Billing 23 finally succeeded in producing a highly
luminous homogeneous hydrogen canal ray by using the new ion-optical
method, and in proving the interference of H-light as emitted from canal
rays of varying places as predicted in Einstein's tilting-mirror experiments.
With experiments conducted under various conditions he also illustrated -
surpassing even Einstein's theory of 1926 - its conformity with Heisenberg's
Uncertainty Principle of 1927. It should also be mentioned that at the same
time Gerhard Otting 24 from our institute and H. J. Ives and G. R. Stievell
in the U.S.A. had proven quantitatively the relativistic quadratic Doppler
effect using transverse observation of homogeneous canal rays. One might
say that such a belated confirmation was not actually of any importance.
To my mind, however, the importance which Einstein himself had attached
to this experiment in 1926 justifies the efforts, not to mention the fact that
every quantitative experiment in physics has its own intrinsic value.
We never found out whether or not Einstein ever received the reprints
of the Annalenarbeit (1938), the credit for which was given him by name.
On the occasion of Albert Einstein's seventieth birthday (March 14,
1949), the Oberbiirgermeister of Ulm, Theodor Pfizer, invited a large audience
to honor the city'S "favourite son". Everyone sat crowded together on fold-
ing chairs and wooden benches in the old "Schuhhaussaal". Pfizer sent a
copy of my speech and his to Princeton. Einstein who at the time was reluc-
tant, even negative in regard to attempts at renewing old acquaintances, ans-
wered him on September 28: "Thank you for sending a copy of the article
telling of your celebration and for your letter of August 17. Please thank my
colleague Mr. Gerlach for the friendly efforts which he undertook upon this

22 W. Gerlach u. E. Riichardt, Uber die Koharenzlange des von Kanalstrahlen emittier-

ten Lichtes, Ann. d. Phys. 24, 1935, S. 124.
23 H. Billing, Ein Interferenzversuch mit dem Lichte eines Kanalstrahles. Ann d. Phy.
32,1938, S. 577.
24 C. Otting, Der quadratische Doppler-Effekt, Phys. ZS 40, 681, 1939. H. I. Ives u.
G. R. Stievell, Journ. Opt. Soc. 28, 215,1938.
200 Walther Gerlach

occasion. We live in a time of tragic and confusing events, so that one is

doubly happy for any sign of human understanding."
"The times are more concerned with bullets and cannonballs than with
the moonball" as Ludwig Kepler had complained three hundred years
before, 1634, when the work of his father, "Astronomia Lunaris" was
published.
 201

Mercer Street and other Memories

gathered by John A. Wheeler

The famous 1950 Japanese film Rashomon recounts a dramatic episode three times
over in the very different versions perceived by three of those who took part. The
account of a May 16, 1953 visit to Einstein given here differs in that there are four
versions supplied respectively by John Wheeler, Marcel Wellner, Arthur Komar and
O. W. Greenberg. The editor (J. A. W.) expresses his appreciation to them for the
permission to quote them, and to the Albert Einstein Estate for permission to
quote brief passages from Einstein's writings.

My first chance to see and hear Albert Einstein came one afternoon in
the academic year 1933-34. I was in my first year of postdoctoral work with
Gregory Breit in New York. He told me that there would be a quiet, small,
unannounced seminar by Einstein that afternoon. We took the train to
Princeton and walked to Fine Hall. Unified field theory was to be the topic,
it became clear, when Einstein entered the room and began to speak. His
English, though a little accented, was beautifully clear and slow. His delivery
was spontaneous and serious with every now and then a touch of humor. I
was not familiar with his subject at that time but I could sense that he had
his doubts about the particular version of unified field theory he was then
discussing. I had been accustomed before this to seminars in physics where
equations were taken up one at a time or, if I may say so, dealt with in retail
trade. Here for the first time I saw equations dealt with wholesale. One
counted the number of unknowns and the number of supplementary condi-
tions and compared them with the number of equations and the number of
coordinate degrees of freedom. The idea was not to solve the equations but
rather to decide whether they possessed a solution and whether it was
unique. It was clear on this first encounter that Einstein was following very
much his own line, independent of the interest in nuclear physics then at
high tide in the United States.
In 1938 I moved to Princeton and at infrequent intervals called on
Einstein at his house at 112 Mercer Street, climbing the stairs to his second
floor study that looked out on the Graduate College. Once discussing with
him my hopes some day to understand radiation damping in terms of the
interaction between the source and the absorber, he told me about his de-
bate with W. Ritz [1]. The two men joined to write up their contrasting
202 gathered by John A. Wheeler

points of view in a joint paper. In it Ritz argued that the elementary interac-
tion is responsible for the irreversibility. In contrast Einstein favored the
view that elementary interactions are time symmetric and that any irreversi-
bility is caused by asymmetry in time of the initial conditions. He also
made reference to a fascinating paper of Tetrode [2] on the same question.
Especially vivid in my mind is a call I made in 1941 to explain the "sum
over histories" approach to quantum mechanics then being developed by
Richard Feynman [3], whom I was fortunate enough to have as a graduate
student. I had gone to see Einstein with the hope to persuade him of the
naturalness of quantum theory when seen in this new light, connected so
closely and so beautifully with the variation principle of classical mechanics.
He listened to me patiently for twenty minutes until I finished. At the end
he repeated that familiar remark of his, "I still cannot believe that the good
Lord plays dice" [4]. And then he went on to add again in his beautifully
slow, clear, well-modulated and humorous way, "Of course I may be wrong;
but perhaps I have earned the right to make my mistakes".
One day something made me say, "Professor Einstein you must often
be invited to other places. Are you never tempted to visit?" "I love to
travel", he replied, "but I hate to arrive".
In the Fall of 1952 I gave for the first time the course in relativity,
general and special, from which I was to learn so much from my students
over the years. On May 16, 1953, not quite two years before he died, Ein-
stein was kind enough to invite me to bring the eight to ten students in the
course around to his house for tea. [The recollections kindly provided by
three of them follow. Arthur Komar, Marcel Wellner and o. W. Greenberg.]
Margot Einstein and Helen Dukas served tea as we sat around the dining
room table. The students asked questions about everything from the nature
of electricity and unified field theory to the expanding universe and his
position on quantum theory and Einstein responded at length and fascinat-
ingly. Finally one student outdid the other in the boldness of his question:
"Professor Einstein, what will become of this house when you are no longer
living?" Einstein's face took on that humorous smile and again he spoke in
that beautiful, slow, slightly accented English that could have been con-
verted immediately into printer's type, "This house will never become a
place of pilgrimage where the pilgrims come to look at the bones of the
saint." And so it is today. The tourist buses drive up. The pilgrims climb
out to photograph the house - but they don't go in.
A further encounter was my last. We persuaded him to give a seminar
to a restricted group [see notes on it below]. In it the quantum was a central
topic. No one can forget how he expressed his discomfort about the role of
the observer. "When a person such as a mouse observes the universe, does
that change the state of the universe?"
Mercer Street and other Memories 203

In all the history of human thought there is no greater dialogue than

that which took place over the years between Niels Bohr and Albert Einstein
about the meaning of the quantum. Their discussion has already been depict-
ed in sculpture and surely will be described some day in pictures and words.
Nobody can forget Einstein's letter to the young Bohr when first he met
him: "I am studying your great works and - when I get stuck anywhere -
now have the pleasure of seeing your friendly young face before me smiling
and explaining" [5). There is no greater monument to the dialogue than
Bohr's summary of it in Paul Arthur Schilpp, editor, Albert Einstein: Philos-
opher-Scientist [6).

References

[1] Ritz, w., and A. Einstein, "Zum gegenwartigen Stand des Strahlungsproblems",
Physik. Zeits. 10, 323-324 (1909).
[2] Tetrode, H., "Ober den Wirkungszusammenhang der Welt. Eine Erweiterung der klas-
sischen Dynamik", Zeits. f. Physik. 10, 317-328 (1922).
[3] Feynman, R. P., "The principle of least action in quantum mechanics", doctoral
dissertation, Princeton University, 1942; unpublished; available from University
Microfilms, Inc., Ann Arbor, Michigan 49106.
[4] Einstein, A., letter to Max Born, December 12,1926.
[5] Einstein, A., letter to Niels Bohr, May 2,1920.
[6] Schilpp, P. A., ed., Albert Einstein: Philosopher-Scientist, Library of Living Philoso-
phers, Evanston, Illinois. 1949, and subsequent paperback editions elsewhere.
204 gathered by John A. Wheeler

Arthur Komar's Remembrances of the May 16, 1953: Visit of Class in Rela-
tivity to Einstein's Residence at 112 Mercer Street, Princeton

as told to John A. Wheeler, August 8, 1977

There were about eight of us. We sat down and had tea. Miss Dukas
brought the tea. Einstein said he was so pleased to have some contact with
young people. John Wheeler asked him about the Einstein-Rosen bridge.
Why did he first introduce it then drop it? Einstein said he originally thought
of it as a unique structure reaching across two nearly flat sheets. However,
when he realized it was not unique, the bridge seemed to him unwieldy,
unattractive, and offering too many possibilities. It was not clear what to do
with them all. Arthur Komar asked what he thought of the idea of Edding-
ton for getting the dimensionless constants. Einstein replied that he was
much interested in finding a theory or an understanding of the dimension-
less constants; but he felt that there was no solution of real interest avail-
able at that time. Komar did not remember questions about the expanding
universe of gravitational radiation or the nature of electricity. He did recall
that vivid phrase at the end, "this house will never become a place of pil-
grimage where the pilgrims come to look at the bones of the saint". The
whole encounter was in English. Komar also remembered Einstein's coming
to Palmer Physical Laboratory and giving a talk containing two striking
comments. (1) The laws of physics should be simple. Someone in the au-
dience asked, "But what if they are not simple?" "Then I would not be
interested in them." (2) Einstein was asked why he rejected quantum me-
chanics. He said he could not accept the concept of a priori probability.
Someone in the audience said, "But you were the one who introduced
a priori probability, in the A and B coefficients." "Yes, I know that and
have regretted it ever since; but when one is doing physics one should not
let one's left hand know what one's right hand is doing." At the end of
this lecture he sat down, leaned back, sighed and said, "This is my last
examination. "
Mercer Street and other Memories 205

Excerpts from a 10 September 1977 Letter from Marcel Wellner to John

A. Wheeler

My memory of our meeting with Einstein [May 16, 1953] is still

fairly vivid. I seem to recall that most of us were too intimidated to ask
him very much, and you had to be our spokesman. This was in spite of
the fact that we were pretty well prepared. I am enclosing [see below]
a copy of a problem assignment which you gave us shortly before that,
and which reflects our preoccupations during the course you were teaching
at the time. Your questions 1 and 2 seem particularly relevant here.
At Einstein's you asked on our behalf what were his thoughts on
Mach's principle. This must have been somewhat removed from his own
concerns at the time, because you had to repeat the word 'Mach' in its
German pronunciation to clarify the question. (As you can see, my memo-
ry is mainly of an auditory kind.)
His answer was somewhat disappointing - at least to me. He said he
no longer held to his earlier views about Mach's principle, and that perhaps
there wasn't in nature anything corresponding to Mach's principle after
all.

Problems
[for class in relativity in May 1953, a few days before the hoped-for visit
with Einstein]

1. Present Mach's principle in a form as eloquent and clear as possible, and

independent of any reference to Einstein's theory. At the end of the
presentation, give a brief summary of the points still to be investigated
in order that this principle should have a satisfactory and logically sound
mathematical formulation.
2. List three questions that you would like to put up to Einstein, with a one
paragraph elaboration of each.
3. Treat the problem of an infinitesimal test particle started from rest from
an arbitrary point in the Schwarzschild field.
4. Derive the Schwarzschild values for a and b in

(dS)2 = a(dr)2 + r2 [(d8)2 + (sin8dl/l)2] + b(dt)2

from the variation principle 8JfJfRd (vol) = o.

206 gathered by John A. Wheeler

Mercer Street

A spring afternoon,
A line of nine walk through the town,
A musty house, the shutters drawn,
A sage lives within.

His key turned the lock

For twenty years, to unify
Electric field, magnetic field,
Space-time, matter, too.

A calm beyond time,

A humble man, received his guests.
To talk, to feel the breath of youth,
To hand them the key.

The day turned to dusk.

The parting time. Advice was sought
For these young men who start the path
He lost long ago.
He shrugged, scratched his head.
Discomforted, at sea, he sent
Them out with "Who am I to say?"
Cool air cleared their heads.

Oscar Wallace Greenberg

Mercer Street and other Memories 207

Last Lecture of Albert Einstein

Room 307, Palmer Physical Laboratory, Princeton University, 14 April 1954

at relativity seminar of J. A. Wheeler. Introduced by O. W. Greenberg.
Notes taken by J. A. W. at the time.

Quantum theory: In which sense is it not final? "Classical" quantum

theory - founded on Hamiltonian equations, similar to electrodynamics
founded on Maxwell equations. How did I become a heretic? Radiation
raises system to a higher state. One can weaken the field indefinitely. The
system is raised more and more rarely to the higher state. The probability
becomes infinitely small to produce a finite effect. One cannot of course
formulate this situation satisfactorily in any mathematical scheme. There-
fore, one is led to a probabilistic description. One finds himself saying that
probability is a definite part of reality. It is an advantage that the law of
Coulomb can be used in the new scheme, by translation from classical theo-
ry. I am a heretic. If radiation causes jumps, it must have a granular character
like matter.
What is really the meaning of 1/1? Can't believe that the state in quan-
tum theory provides a complete representation of the physical situation.
Consider a sphere 1 mm in diameter. One can see it with the eye. It can go
to and fro between two planes, ideally elastic. One can forget the internal
coordinates of the sphere. Consider a state of fixed energy (Fig. 1). If one
neglects the fine structure, all places have the same probability. More accu-
rately, there are some places where the thing can never be. This is contrary
to the ordinary Newtonian idea of motion. There is no question that this is
true; it certainly corresponds to reality. Fourier analysis shows there is a
probability of 1/2 for v = vo , and 1/2 for v = - Vo.
lt is difficult to believe that this description is complete. It seems to
make the world quite nebulous unless somebody, like a mouse, is looking
at it. The problem is to understand that one can observe the particle with a
lantern.

Fig. 1
"Consider a sphere lmm in
diameter. It can go to and
fro between two planes,
ideally elastic."
208 gathered by John A. Wheeler

The scheme is of very great practical value as long as we have nothing

better, and makes good use of the concepts such as mass and charge with
which physics started in the earliest days. But one has to mistrust it if one
believes in a deeper scheme.
The Maxwellian scheme is marvellously effective in explaining many
things, particularly macroscopic. But it runs into trouble on radiation.
Fluctuations are bigger on Planck's law than they are on Maxwellian theory.
I knew in constructing special relativiry that it was not complete. So is
everything that we do in our time: with one hand we believe; with the other,
we doubt. I once thought temperature a basic concept. I feel the same way
about Maxwellian theory. But I am now convinced there is no cheap way
out. If there are too many hypothetical elements one cannot believe one
is on the right track. Thus I came to logical simplicity, a desperate [man's]
way to get on the right track. But one event in my life convinced me of the
usefulness of logical simplicity. That was general relativity.
It can be looked at as a theory which makes us independent of the
inertial system. The concept of inertial system was regarded by the founder
himself, and his scientific enemy, Huygens, and Leibniz, as exceedingly
unclear. For Galileo, acceleration is the fundamental concept on which
mechanics is founded. But what is acceleration? Newton invented the
infinitesimal calculus. But this really doesn't provide an answer. There
are coordinate systems that are inertial and others that are not. A coor-
dinate system is satisfactory if in it the equations of motion hold. In classi-
cal theory there are three independent concepts: space, time, material
points. The behavior of material points is determined by the inertial system.
But this is like God Almighty, unaffected by anything else. Newton recog-
nized very clearly it was very hard to regard space as something absolute.
This is not the direct way I found the theory of gravitation. The real way
is a very strange story.
I had to write a paper about the content of special relativity. Then
I came to the question how to handle gravity. The object falls with a differ-
ent acceleration if it is moving than if it is not moving (Fig. 2). Thus a gas

1
Fig. 2
"The objects falls with a different
acceleration if it is moving than if
it is not moving."
Mercer Street and other Memories 209

falls with another acceleration if heated than if not heated. I felt this is
not true. Came out that acceleration is independent of quality of matter:
pendulum experiments.
Change coordinate system? Then change acceleration. Then I came to
a real understanding of the equivalence of gravitational and inertial mass.
No inertial system can be preferred. That was not so clear to me at that
time. But Mach had the same idea, not the relation of gravitational and
inertial mass, but that "inertial system" was a very vicious concept. Inertia
come from the presence of other bodies? How possible? Relative accelera-
tion and against this a resistance. Quite a nice idea. But if you give up space,
you have an enormous number of distances, and unhandy consistency rela-
tions. Mach not aware of time concept. Great thing that Mach, centuries
after Newton, felt that there was something important about this concept
of avoiding an inertial system. Need absolute covariance.
Not yet so clear in Riemann's concept of space. His curvature is abso-
lutely covariant. But this was not so clear at the time this work was present-
ed by Riemann and his successors. The first to see this clearly was Levi-
Civita: absolute parallelism and a way to differentiate. Recognized that
possibility to avoid inertial system depended on existence of a r-field that
described parallelism in the infinitesimally small. This was a great advance.
It made it possible to see how to generalize relativity to include electromag-
netism.
The representation of matter by a tensor was only a fill-in to make
it possible to do something temporarily, a wooden nose in a snow man.
The theory wasn't complete, because we know the world is not limited
to gravitation. After decades I came to the idea to generalize by using
unsymmetric gij'S, the method of logical consistency gone mad. I was very
conscious of this objection. r is a field. Without this there is no hope to
express things in general relativity.
Present quantum theory based on special relativity is horribly compli-
cated. For most people special relativity, electromagnetism and gravitation
are unimportant, to be added in at the end after everything else has been
done. On the contrary, we have to take them into account from the be-
ginning. Otherwise it is as if one did a classical problem and put in the law
of conservation of energy at the end. Expect to describe a system only by
quantum numbers. There is much reasonable in this. But a field theory
seems to present us with an infinite number of quantum numbers. There
is much reason to be attracted to a theory with no space, no time. But no-
body has an idea how to build it up. Of course, to quantize space and time
is a childish idea. This is my excuse for feeling so strongly. It is pedagogic
to insist that if one has a field theory, one must demand solutions without
singularity. If a singularity is allowed, there are too many arbitrary assump-
tions, and too much arbitrariness.
210 gathered by John A. Wheeler

Questions:

Greenberg: Do you have a different interpretation of deBroglie waves?

Einstein: I consider them almost comparable in reality to light waves,
but not quite. There are so many fields as there are masses;
and then the trouble of higher ...... [word missed].
Callaway: The equivalence principle says it is reasonable to absorb gravi-
tation in the metric. But no such principle is known for elec-
tromagnetic theory.
Einstein: I believe that the concept of motion has no place in a unified
theory; that "geodesic lines" are only a provisory concept, a
stop-gap. Also it is against the whole idea of quantum theory.
Greenberg: Is there anything in gravitation theory corresponding to radia-
tion?
Einstein: This is a headache. Why not expect also gravitational quanta?
But this is hard to include in field theory. Perhaps this is an
objection to any field theory that has to do with gravitation.
Electromagnetic waves can be put into a container, but gravi-
tational waves cannot - there are two signs of charge for the
one; only one for the other. Thus there is a difference in kind
between the two theories. This gives the feeling that gravita-
tion is not more true than any classical theory. There is an
infinity of constants in a field theory. Only way to overcome:
if the condition of nonexistence of singularities 10 a very
unenigmatic way fixed up quantum numbers.
Carlson: What do you think of Bohm's theory?
Einstein: It is clever, but I don't believe it. It is outrageous to believe
that the particle between walls does not move.
Komar: Why is unified theory simpler than projective theory?
Einstein: The most basic thing in relativity is the replacement of a field
by something like a r. In gravitation we have to symmetrize.
So we can think unified theory is a simplification of gravitation
theory. But gravitation theory is a field theory, and may have
difficulties.
Callaway: What do you think of Mach's principle?
Einstein: Mach assumes matter may be permanent. Therefore, why sepa-
rate the rest of the field from the gravitational field? We have a
satisfactory description in conformity with Mach's principle
if we have a theory without boundary conditions. I once thought
Mercer Street and other Memories 211

to take the universe roughly static and closed in space. Then

came the cosmological A.. But it was a sin against mathemati-
cal simplicity. If the world is expanding, it is hopeless. Time
is essential. A boundary is unthinkable. Your question is related
to the role of matter.
Mozeley: Must a field theory be deterministic?
Einstein: [first few words missed] It is the negative part of determinism
that probability should not intervene, because it is not a quali-
ty of the system. [End.]
 213

Einstein - and the Vanity of Academia

Wolfgang Yourgrau

It is a labour of love to put down, not methodically, not genuflecting,

but in a desultory, vague, rambling mode a few reminiscences of the most
consummate, momentous thinker this age of unparalleled advance in theore-
tical physics, astrophysics, and biophysics can boast of. The reader will
hopefully, if not graciously, forgive me for not cumbering him in these
personal reflections with technical excursions. The seminal role Einstein
played and still occupies in the most crucial domains of physical reasoning
is commonplace. Besides, any form of hagiography, of idolatry had always
been repulsive to him. Hence, I shall not resort to hyperbole or worse still,
to an apotheosis. Be it as it may, I am confident that "Dieu me pardonnera
- c 'est Son metier. "
I met him first as a student, in 1927 or 1928. We physics students were
of course well acquainted wi,th his legendary frame. Even in night clubs, in
cabarets comedians referred, often in a tasteless manner, to this "quixotic,"
"abstruse" professor who looked more like a poor Italian musician, like the
image of a poor fiddler than like a dignified, staid German professor. Need-
less to stress that we were not in a position to do justice to his lectures, to
appreciate the originality, creativity, and boldness of his theories, Moreover,
his Ulmer dialect was very marked and amused the "purists" among us few
attendants. He lectured in a soft, almost subdued voice; his lecturing techni-
que - in contrast to that of Schrodinger, Pringsheim, von Laue, E. Schmidt,
et al. - was not impressive at all.
In point of fact, Einstein was not an inspiring, not a stimulating acade-
mic teacher. He did not evoke enthusiasm among us neophytes. None the
less, although he did not possess the gift of the gab, rarely indulged in melli-
fluous metaphors, we strongly felt that he spoke "canonically," as a nun,
a student, expressed her enchantment with that great man. Those big, sad
eyes, the brows raised like those of a melancholy clown; they will never be
forgotten by all who knew him then.
Most of us found it difficult to absorb the very meaning, not to men-
tion the significance, of his papers and books. If the covers of some books
in physics are much too far apart, Einstein's written works were however
too compressed, too exacting, too abstract for our liking. The extrapolative
distance between the local and the global was rendered more enormous,
214 Wolfgang Yourgrau

less intelligible, because he frequently, no, nearly always, omitted simple,

explicit definitions. No wonder, then, that in our frustration we consulted
von Laue who had a special talent for unequivocal, unambiguous definitions
and explanations. Once or twice, some of us asked Einstein after the lecture
about the connotation of a particular term or relation. He smiled benignly -
'ecumenical' would describe his countenance more accurately - and very
patiently attempted to 'enlighten' us, the still exoteric disciples. The results
were frustrating: we left the room totally bewildered, unable to fill the inter-
polative gaps between the theorems or propositions in question.
Now, that I am no longer a neophyte in the temple of physics, I realize
that the uninitiates' quest for explicit definitions is no cogent condition for
the understanding of a physical (or mathematical) hypothesis or theory.
Someone discovered the following definition of the harmless term 'net': a
"series of reticulated interstices." Since I read this monstrous, recondite
definition, my esteem for Hilbert (and hence for Einstein) has steadily grown.
It was Hilbert who refrained from defining - neither explicitly nor opera-
tionally - entities or quantities such as points, squares, and cubes. Einstein
followed the example of Hilbert. Yet Einstein's writing and teaching contain-
ed an additional, rather vexing, characteristic. After a very complex equa-
tion, full of physically or cosmologically highly sophisticated terms, identi-
ties and/or inequalities, he writes: " ... where c is the velocity of light."
Many an author regrettably parrots the master too religiously in this respect.
In the early thirties I experienced a serious railway accident. After
having visited my father and my stepmother in Brussels (my father was Bel-
gian, my mother German), a fire broke out in the night train to Berlin.
For miles and miles the train had races through Belgium and to a part of
Germany like a burning torch. Passengers suffered carbon monoxide poison-
ing and various bodily injuries. I had jumped, almost unconscious, through
the window and finally landed in the Gewerbekrankenhaus in Berlin, where
I was treated by the director, professor Baader, for 4-1/2 months.
Einstein had learnt from Dr. Dinkin, a Russian refugee physician who
knew Einstein well, and also from an uncle of mine, also a physician, that
I had to be hospitalized due to an accident. My uncle lived about 100 yards
from the Einsteins' flat in the HaberlandstrafSe. On occasion, when Einstein's
personal physician was not available, my uncle had treated the great man.
Well, incredible as it may sound, Einstein alone or with his kindly, conside-
rate wife, visited me at least twice a week, until I was allowed to leave the
hospital.
The two sat a few feet away from my bed on uncomfortable chairs.
He was shy, self-conscious. My head, arms, and legs were heavily bandaged, I
felt dizzy, double vision set in sporadically, and so did vomiting. Frau Pro-
fessor displayed genuine maternal warmth, reported regularly my respective
state of health to my mother and even wrote to my father in Brussels. No
Einstein - and the Vanity of Academia 215

relative, no friend, no colleague (s) showed to me so much consideration and

earnest concern than the Einsteins.
Frau Professor brought me plenty of delicious fonds which, to my dis-
may, I was of course neither allowed to have nor capable of eating. Our
conversation was onesided by reason of my injuries and due to physical and
mental exhaustion. Yet I recall vividly that he enthused about the genius of
Mozart, the impact of Galileo upon relativity. Einstein also wanted to be
nice. And so he told me how fortunate I was that Schrodinger had taken a
great liking to me. He had learnt from Professor Baader that I could not
utter a single word or sound until I had reached to age of 3 112 years. To
console me, the great thinker confided to me in his soft voice that he too
could not talk at all before the age of 3. I felt quite elated that at least this
one trait, though a defect, I was 'privileged' to share with him. Strangely
enough, he also knew that for many years I had been an incurable stutterer
and went through agony and shame when I had to give a talk in our seminar.
He assured me that I had markedly improved, and that stuttering or stam-
mering was an ephemeral phenomenon, much easier to cure than a lisp.
In an attempt to "cheer me up," Einstein advised me to study Helm-
holtz intensively and gave me 2 booklets by Helmholtz as gifts. The Ein-
steins left me after 30-40 minutes. On leaving, he always put a closed
envelope on my nightstand. it contained 10 Mark and a few lines saying
that I should buy myself a book or anything else I fancy.
I must have received no less than about 160 Mark from him, in postal
orders. How embarrassed I was! The reason? My parents, divorced for many
years, vied for my affection by bribing me with expensive presents. It was
common knowledge in the department that Einstein was the highest paid
theoretical physicist in Germany; Planck had insisted upon it. But my
father, a well-to-do organic, independent chemist and my mother likewise
possessing independent means, were so irrationally attached to me, their
only child, their "Bubi," that I lived like a foppish snob in an elegant 3
rooms, had a man servant (I), owned elegant clothes and a Fiat.
In brief, the postal orders of Einstein made me feel rather sheepish.
What could I say? Return the money and thus perhaps hurt him and his
wife? What a quandary I found myself in! Had I not been spending more
than 10 Mark as a tip on some rather frivolous occasions? Had I not collect-
ed a huge number of books on physics, mathematics, logic, music, litera-
ture etc.? Later, colleagues told me what a fool I was for having exchanged
those postal orders into cash and bought beaucoup de petits cadeaux for my
nurses and physicians. I should have kept those tokens of kindness by that
intellectual giant for posterity, or at least for my future children.
Some time in 1929 or 1930 I received a very generous research grant
from the Lincoln Foundation; it was renewed for another year. The British
mountaineer who had climbed that Mont Blanc twice, once with a wooden
216 Wolfgang Yourgrau

leg - his name was G. W. Young - had approached Einstein, von Laue, and
particularly Schrodinger, whose assistant I was, as to whether I deserved such
a big grant for my research project. I was lucky and quite proud of myself.
I later learnt that Einstein's recommendation was the briefest one. All this
took place without my knowledge. According to Lao Tse:

Those who know, do not tell;

Those who tell, do not know.

Einstein had no guile, no fanged tongue, and he never punditized. None

of us students could ever impute to him arrogance, vanity, or conceit. A
Spanish proverb states that a man is as great as the problems that irritate
him. Einstein was very fortunate indeed that he was, perhaps constitutional-
ly, incapable of being disturbed or affected by petty issues, by unfounded
criticism, or by flattery of any kind. I remember only one exception, namely,
his frequent disagreements with Nernst. In June, 1913, Planck, Nernst,
Rubens, and Warburg proposed Einstein to become an ordentliches Mitglied
der Akademie with the enormous salary of 12.000 Mark. Einstein was of
course not unaware that Nernst had avidly supported Planck concerning Ein-
steins receiving this highest distinction for a scholar, next to the Nobel
award.
I speak from bitter personal experience that the excellent scientist
Walther Nernst was a vain man, and displayed some of the most objec-
tionable features of a Teutonic professor. I was not the only one among the
students, assistants, and professors who shared my strong antipathy to the
"father of the NERNST law." (He himself employed this phrase in his terrib-
ly boring, ego-centric lectures.) Now, it was known to us that Einstein did
not like Nernst - for reasons never made public.
And so it is with some malicious glee that I pass on to the reader an
anecdote told to me by Schrodinger. In a meeting of the physical society
in Berlin, Nernst pontificated and did not even tolerate the views of some
other speakers because they differed fundamentally from his own doctrine.
In other words, it never entered his mind that both, his opponents as well as
he himself, may be wrong.
Einstein did not concur with Nernst's arguments at all, but refrained
from any polemic. In the end, however, Einstein's patience snapped like a
string on a violin. His comments were pithy and inexorably to the point.
"I wonder whether Herr Nernst will ever be able to fathom the depth of his
ignorance." First there was silence. Then the crowd burst unashamedly into
loud, hearty laughter. It is possible that this anecdote is spurious, invented
much later by Schrodinger who had no particular sympathy for Nernst; in-
deed, for none of the "Nernsts" populating the universe of physicists or
physical chemists. Schrodinger was a man of imagination, an artiste manque
- to wit, his beautiful little volume containing passionate love poems. In
Einstein - and the Vanity of Academia 217

one of his seminars, he quoted Weierstrass' very un-Nernst-like remark: "A

mathematician who is not also something of a poet will never be a complete
mathematician." I therefore merely repeat Einstein's sarcastic remark on
Nernst, as told to me by Schrodinger, without claiming its veracity. No
doubt, Schrodinger could have let his imagination, his creative phantasy
run away with him. He might have either borrowed that phrase or even
coined it.
My mother and I were passionately addicted to chamber music. One of
the finest conductors of a then highly acclaimed chamber orchestra was
Michael Taube. Mother and I had our regular seats in the 4th and 5th row.
In the middle of the concert (I remember it well: it was one of Handel's
concerti grossi), my mother pointed indignantly at a man who sat about
6 seats to the right, in the row in front of us. With her customary, i. e.,
stentorian alto she turned to me: "Who is that fellow with the long, un-
kempt mane and a collar covered with dandruff -like snow?"
In vain did I try to draw her attention to the music. Mother was educat-
ed, well-read, a good chess player and inordinately musical - but she was
fascinated by that man in the row in front of us. To keep her quiet, I whis-
pered - in a low voice trembling with awe and anxiety -: " ... this is Ein-
stein, THE Einstein, one of my most esteemed professors, a genius, one of
the greatest -." It did not help. Without being impressed, she gave once
more vent to her indignation, in her piercingly sonorous voice to intran-
signent disapproval.
During the interva\ I claimed intolerable headache and left. To this day
I do not know whether Einstein had noticed our little "dialogue" behind
his back.
An uncle of mine married a niece (or more remote relative) of Einstein.
He was editor of a large newspaper. The Einsteins sent a precious wedding
present, but were unable to attend the wedding ceremony which took place
in Munich. Two years later that marriage broke up - due to his being an in-
corrigible philanderer. Einstein met him afterwards quite frequently on seve-
ral occasions. Never did he show any disapproval or apprehension. My fickle
uncle expected to be slighted or reproached. Nothing of that sort ever
happened; both Einsteins continued to treat my irresponsible uncle in a
friendly, not just correct manner.
In Faust I, man is described as "microscopic fool." Geothe certainly
did not anticipate a "macroscopic fool" like Einstein. The word 'fool' has
many connotations - some are even tantamount to praise, e.g., in the
middle ages or in the work of Hermann Hesse. During the last 16 years I
did some research work with my very dear friend H.-J. Treder, perhaps
(next to Peter Bergmann) the most devoted and erudite Einstein scholar.
Recently Hans-Jiirgen showed me an item in an old issue of Annalen der Phy-
sik. It was too funny for words: Here Einstein, in his laconic style, informed
the reader that he had made a few mathematical mistakes in his doctoral
218 Wolfgang Yourgrau

dissertation accepted in toto, many years before these published errata. The
joke is, of course, on his examinators; the 'macroscopic' fool-hero is the man
who already in his late forties looked like an Italian virtuoso - some even
compared his appearance to that of a pavement artist.
This little item in the Annalen speaks more eloquently for the nobility,
the integrity of that unique giant in physics than any a paean sung in honour
of this homme celebre.
Many weird experiences with Einstein the academic teacher come to my
mind, now that I try to recall the remote past. We students often grumbled
that he had "intercourse with the blackboard" - as we put it facetiously.
He wrote and wrote and wrote, without realizing as to whether the content
of his statements would make sense to us. Often times he stopped in the
middle of writing down a relation or proposition or theorem, became lost in
thought, for minutes, and then produced some entirely novel hypothesis or
theory whose dramatic ingenuity even we ignoramuses recognized as 'com-
mensurate with the explosion of a supernova'! In those moments we were
not aware of his many foibles, his clumsiness, his lack of rapport with the
majority of the class, his oddities, his frequently almost inaudible delivery.
Only later did I learn that he had developed into a fine lecturer.
The reader will be lenient with my crime de lese majeste - I was too
young to appreciate all the virtues Einstein exhibited already in those days
to those who had an open mind. But I was a Schrodinger man, whose tem-
perament, elegance of exposition, humor, and charm, apart from his genius
as a physicist, had me completely captivated. I even imitated his gait, dressed
like him, adopted his mannerisms, and became a ludicrous copy of him, who
loathed hero-worship. He treated me with acid sarcasm when he perceived
my helpless parody - in fact, he accused me of travesty, of caricaturing
him. Still, I became his assistant and later his younger friend. I soon learnt
that Schrodinger had nothing but highest admiration for Einstein, as a
scholar, a human being, a model of unblemished moral conduct and recti-
tude.
When Friedrich Schiller's centennial of his death was the object of end-
less memorial celebrations by the press, the public, the literati, the painter-
poet Morgenstern decided to convert the name of the beloved titanic genius
into Max Zottuk. All this feverish glorification, this puerile apotheosis was
just too much for him who loved and admired Schiller in stillness, and with
passion. Well, I sympathize with Morgenstern and would prefer to call Ein-
stein no longer by his name, but toy with the idea to dub him Berthold Wal-
dinger. B. W. could have afforded to neglect his external appearance, social
duties, academic conventions. He would have overcome his innate aversion
to close relations with people. The epithets "egghead," "odd ball," "enfant
terrible" would have never aroused his ire - at most, his subtle sense of
humour.
Einstein - and the Vanity of Academia 219

Berthold Waldinger alias Einstein would have enjoyed some apparent

abstruse consequences of his theories with puckish delight. Once Einstein
alias Waldinger was asked to explain the quintessence of RT in a few words.
Without thinking he obliged:

It was formerly believed that, if all material things disappeared from the universe,
time and space would be left. According to RT, however, time and space disappear
together with the things. This is all.

Born was wrong in propounding that "Nature's attempt to produce a

thinking creature has failed." Waldinger (alias Einstein) is one of the most
convincing counter-examples.
The celebrated Albert Einstein - not Waldinger - is supposed to have
been a good violinist; he also played the piano and the organ. Here we en-
counter a collision between myth and hard fact. Twice I had to play the cello
in a quartet that met on Friday evenings in the Pragerstrage, not more than
100-130 yards away from Einstein's flat. True, the revered master was
extremely musical, had a profound love for Mozart, also for Bach, though to
a lesser degree. Yet his enthusiasm for music, his admirable knowledge of
musicology, must not be confounded with his proficiency as a violinist. I
was quite relieved when the regular Friday-evening cellist took over again.
Having been trained by one of Europe's finest cellists since I was 6 (practi-
cing 2-3 hours daily), I felt it almost degrading to play in that quartet
consisting of Einstein, my uncle the physician, and an old lady who used
to be a piano teacher in a girls' school. The only sound and technically
competent member was the cellist - a pharmacist by profession. My uncle
played the second violin or the viola - both poorly. I can only hope that
the Muse Euterpe took our enthusiasm into account. Still, Einstein played
evidently better than 'his' alias ....
He was a physicist, not a philosopher or mathematician. This was
Bertrand Russell's and Karl Popper's candid opinion. Einstein said on one
occasIOn:

"Since the mathematicians have invaded the theory of relativity, I do

not understand it myself any more."

The word 'theory' originally signified "passionate sympathetic contem-

plation." Only gradually did this passionate contemplation become more
intellectual, more abstract. His creation still betrays something of the origi-
nal meaning of the word. And yet, perhaps Hilbert was right " ... dag die
Kultur der Mathematik von der theoretischen Physik herkommt."
Albert Einstein - in contrast to Schrodinger, with whom I felt greater
affinity - had the soul of a child, of a jester, and a saint. However, he re-
220 Wolfgang Yourgrau

sembled the Viennese Schrodinger in that both were Bohemians by nature,

not 'professional' in their approach to science, kind-hearted, not cut out for
team-work. Neither belonged wholeheartedly to any country, religion, or
even to his own family. They were citizens of the world and perhaps there-
fore aloof in their dealings with ordinary earthlings.
Yet Einstein was shy, avoided the limelight, never cared for the forma-
lities, the traditions, the respectability of 'alma mater'. According to Nietz-
sche, only scoundrels are modest. He was wrong: Einstein's proverbial humi-
lity repudiates Nietzsche's cynicism. Schrodinger could be arrogant. He
could also play the rules of academia, though often with tongue in cheek.
Einstein - NOT Waldinger!- was indeed the "macroscopic fool," he created
his own universe. 'And the angels will weep for him .... ' Does it matter
that he possessed some bizarre, quixotic character traits, too? We physicists
are somewhat relieved that there exists, most probably, but one universe.
Will there always be one Einstein alone?