a »

Science and Its Conceptual Foundations O uantum

- : Dialogue
A series edited by David L. Hull

The shakingof Revolution

Mara Beller

THE UNIVERSITY OF GHICAGO PRESS

Chicago& London
 CONTENTS
e-F

Thinking and discourse are the same thing. ... What wecall i
thinkingis precisely the inward dialogue.
Plato

The roads by which menarriveattheir insights . . . seem to me

almost as worthy of wonderas these matters in themselves.
List of Illustrations xi
Johannes Kepler Preface and Acknowledgments xiii
Novelty and Dogma
Dialogical Creativity 1.’
Revolutions are true as movements andfalse as regimes. Rhetorical Strategies 10
Maurice Merleau-Ponty ; PARTONE DIALOGICAL EMERGENCE
2. Matrix Theory in Flux - 17
Introduction 17
A Revision of the Origins of the Matrix Theory

TA
The Emotional Confrontation between the Matrix
Physicists and Schrédinger 30 |
Born’s Probabilistic Interpretation: A Case Study
“Concepts in Flux” 39
3; Quantum Philosophy in Flux 51
Introduction BL
Positivism in Flux 52
Indeterminism in Flux 59
4. The Dialogical Emergence of Heisenberg’s
Uncertainty Paper 65
Introduction 65
Dialogue with Schrédinger 67 -
Dialoguewith Pauli 79
Dialogue with Dirac 85
Dialogue withJordan 91
Dialogues with “Lesser” Scientists 96
 Contents Contents ix

The Polyphony of Heisenberg’s Uncertainty Paper 103 9, The Copenhagen Dogma: The Rhetoric of Finality
Introduction 103 and Inevitability 191
The Polyphonyof the Notion of Interpretation 106 Introduction 191
The Contingency of Acausality 110 Acausality and the Indispensability of Classical
Anschaulichkeit and the Status of Classical Concepts 194
Concepts 113 Operationalism: From Consistency to Inevitability
The Dialogical Birth of Bohr’s Complementarity 117 Arguments 199
Introduction L17. Bohm on Classical versus Quantum Concepts and on
Dialogue with Schrédinger: The Structure of Indeterminism -. 206
Atoms 122 ‘ 10. Constructing the Orthodox Narrative 211
Dialogue with Einstein and Compton. 131 Introduction: “Whiggish” Historyand “Winner's”

Demanttgngey
Dialogue with Campbell . 135 Strategies 211
Clash with Heisenberg: Setting the Historical Record Discontinuities and Quantum Jumps 214
Straight 138 Indeterminism and Historiographical Doubts 220
Confrontation with Pauli:. 141 11. The Myth of Wave-Particle Complementarity 223

A YERNMSSMAEE
Conclusion 143 Introduction: The Dramatic Historical —
The Challenge of Einstein-Podolsky-Rosen and Narrative 223
the Two Voices of Bohr’s Response 145 Mathematical Physicists and1 the WaveBavtivte
Two Voices in Bohr’s Responseto Einstein-Podolsky- Dilemma 227 -
Rosen 145 Ambiguity and the Wave-Particle Issue 232
Bohr’s Victory? 151 Ideological and Pedagogical Uses of Wave-Particle
Disturbance, Reality, and Acausality 155 Complementarity 237
Bohr’s Doctrine ofthe Indispensability of Classical 12. Complementarity as Metaphor 243
Concepts and the Correspondence Principle 160 Introduction 243 .
The Web of Correspondences and Harmonies 248
PART TWO RHETORICAL CONSOLIDATION “Wholeness” as Metaphor 252
The Polyphonyof the Copenhagen Interpretation Bohr: Mathematics and Common Language 259
and the Rhetoric of Antirealism 171 Metaphorical Appeal and Conclusion 262
Introduction 171 13. Hero Worship, Construction of Paradigms, and
WhatScientists “Need Not” and “Must Opposition 269
Not”Do 173 Introduction 269
The Appeal of Antirealism: Some General Bohr and Hero Worship —.270
Considerations 176 The Issue of Consistency 275
Reality, Classical Concepts, and Symbols —-179 Opposition, Paradigms, and Past Science 276
The Appeal of Antirealism: Bohr’s Version 184 14. Dialogues or Paradigms? 287
Antirealism and Opposition 185 Introduction 287
The Appearance of Consensus and Conclusion 187 Heisenberg’s ‘Closed Theories” and Kuhnian
“Paradigms” 288
Where Did Kuhnian Incommensurability Come
From? 290
| x Contents

Hanson's Incommensurability and the Copenhagen ILLUSTRATIONS

Dogma 294
Sy
Paradigms andthe History of Science 300
Paradigms and Holism 302
Paradigmsand Creativity 304
15. Dialogical Philosophy and Historiography:
A Tentative Outline ‘ 307
In Praise of Disagreement 308
The Philosophical and Historiographical Advantages
of Dialogism 313
Theory as Practice: Between Tools and
Metaphors 317
Truth and Beauty 320
Diagrams
References 327
Index / 355
Heisenberg’s analysis of a y-ray microscope 72
Bohr’s thought experiments in termsofclassical concepts 148

Photographs(following page 190)

Werner Heisenberg and Niels Bohr

Niels Bohr and Albert Einstein
Otto Stern, Wolfgang Pauli, and Werner Heisenberg -
Paul Dirac and Werner Heisenberg
Niels Bohr, James Franck, and Hans Hansen
Erwin Schrédinger, the king of Sweden, and Werner Heisenberg
Walther Meissner, Max Born, and James Franck

Niels Bohr and Wolfgang Pauli

Albert Einstein and Wolfgang Pauli
Frits Zernike
William Duane
Niels Bohr, Wolfgang Pauli, Lothar Nordheim, and Leon Rosenfeld
 PREFACE AND
ACKNOWLEDGMENTS
Ey

This bookis aboutscientific dialogues that made the quantum revolu-'

tion. The description ofthe intricate flux of these dialogues reveals.the
dynamicandintensely personal nature of scientific theorizing. During
the work on this book I came to the rewardingrealization that an on-
going responsivenessto the concerns of others, in addition to being a
basic human value,is a precondition of scientific creativity. This book
deals with the anatomy of scientific discovery (part 1) and with
the strategies of consolidation of the orthodoxquantum philosophy
(part 2). aa! mee :
Myapproach is based on both admirationfor the creation ofscientific
novelty and criticism toward the establishment of scientific dogmas.
This book offers a thorough historical reevaluation of the emergence
and consolidation of the Copenhagen interpretation of quantum phys--
ics—the dominantinterpretation since its launching in 1927 by Niels
Bohr, Werner Heisenberg, and Max Born.
It is not a historian’stask to offer a specific alternative to the orthodox
Copenhageninterpretation. I take no stand on the existing alternatives
to the orthodox philosophy. This book does not deal with the extensive
and lively contemporary research on the philosophical problems of
quantum physics. Rather, my historical, philosophical, and sociological
analysis of the Copenhagen philosophy demonstrates the possibility
andthe needofa viable alternative to the orthodox interpretation.
- As my work on this book progressed, my revision of the history of
the quantum revolution gradually evolved into a generalcritique of the
revolutionary narratives for the description of the scientific change. Fo-
cusing on the quantum revolution, this book provides an analysis of
how revolutionary stories in history of science are constructed, how
division between “winners” and “losers” is fabricated, how the oppo-
sition is misrepresented and delegitimized, and howtheillusion of the
existence of a paradigmatic consensus amongparticipantsis achieved.
xiv Preface and Acknowledgments Preface and Acknowledgments xv

The bulk of my critical analysis of the Copenhagen philosophy of National Endowmentfor Humanities (grant number FA 31327-92) dur-
complementarity and indeterminism is found in part 2, which is more ing my work onparts of this book. I am also grateful to David and
accessible to the general reader and can be read separately from the rest Frances Dibner and the Dibner Foundation for the endowmentof the
of the book. Nevertheless, parts 1 and 2 are mutually illuminating and Bawpara Druss Dibner Chair in History and Philosophy of Scienceat the
conceptually interconnected. The introductory chapter 1 offers an over- Hebrew University of Jerusalem, which I have the honorto hold.
view of the major themesof this book, while the concluding chapter 15 Part of chapter 7 is based on myjoint paper with Arthur Fine (Beller
is an argumentfor the dialogical approach for studying the advance of ~ and Fine 1994). Portions of chapters 2, 4, 6, 8, 9, 13, and 14 are basedin
scientific knowledge. part on my previously publishedarticles (Beller 1983, 1985, 1990, 1992a,
This book waslong in the making, and I have accumulated:a longlist - 1996a, 1997a, and 1997b).
of debts. There are a few scholars to whom my debt and gratitude are . lam grateful to John Sanders for permission to quote materials from
especially deep. Manyideas in this book were developed while I imag- - his collection (1987) of Bohr’s published papers and unpublished manu-
ined addressing Arthur Fine. Arthur’s writings and personal encour- scripts. The originals of Bohr’s unpublished manuscripts are deposited
agement were invaluable. I owe a great debt to Jim Cushing, for his - in the Niels Bohr Archive in Copenhagen. Theyare available in micro-
unbelievably meticulous and caring reading of the draft of this book, film in the Archive for the History of Quantum Physics (AHQP), copies
for his continuous support, and for his important work onrelatedis- of which are deposited in several universities throughout the world.
sues. Yemima Ben-Menahem,a colleague and friend, provided valuable Tam grateful to Felicity Pors from the Niels Bohr Archive in Copen-
criticism with a rare combination of philosophical wisdom andgraceful hagenfor her assistance. I also want to express my gratitude totheli-
encouragement. Sam Schweber andJohn Stachel, who took an early in- brarians of the Van LeerInstitute, Edelstein Library, and the Niels Bohr
terest in my work, were always very generous with their intellectual. Library in theAmerican Institute of Physics for their help. I also grate-
and personal support. Tom Ryckmanattentively read the draft of this © fully acknowledge the courtesy of the Emilio Segré Visual Archives in
book and provided valuable comments. I am grateful to Allan Franklin ' the AmericanInstitute of Physics for the photographs appearingin this
for many stimulating and pleasant conversations we had during my book, and I thank Jack Scott, the photo administrator, for his helpful
stay in Boulder. ; assistance.
I am fortunate to have a friendly and ‘stimulating environment, with Very special thanks are due to my friend Shuli Barzilai, who returned
Yehuda Elkana, a teacher and a founder of an extensive academic ac- two years ago from hertrip to Russia with a unique souvenir—the
tivity in history and philosophyof science in Israel, and with colleagues photo she took ofthe statue of Einstein and Bohr in a park in Moscow.
at The Hebrew University of Jerusalem such as Yaron Ezrahi, Rafi Falk, The photo decoratesthe jacketof this book.
Michael Head, Itamar Pitowsky, Mark Steiner, and Isachar Unna. Iam one of those lucky authors who has had the pleasure and the
Mylist of debts is a long one, from Stephen Brush, my Ph.D.adviser privilege of working with Susan Abrams, executive editor at the Uni-
in the early days of my academiclife; to Shelly Goldstein and Anna versity of Chicago Press, and to benefit from her sharp mind and big
Sfard, with whom recently engaged in fruitful dialogues. I am also heart. I am grateful to the wonderfulstaff at the University of Chicago
grateful to the following people fortheir stimulation, support, or assis- Press: to Rodney Powell and Charles Clifton, editorial associates for the
tance at some stage of working on this project: Pnina Abir-Am, Gid- sciences; to Leslie Keros, who worked on the proofs; to Martin Hertzel,
eon Akavia, Diana Barkan, the late Yosef Ben-David, Michel Bitbol, who designed the book; to Joan Davies, who supervised the produc-
Jed Buchwald, Catherine Chevalley, Robert Cohen, Olivier Darrigol, tion; and to DavidAftandilian, who handled the promotion. I am also
Alon Drory, Detleff Diirr, Gideon Freudenthal, Michael Friedman, grateful to Diana Gillooly for her meticulous and dedicated work on
James Fuchs, Ruthi Glasner, Galina Granek-Tiroshi, Gerald Holton, the manuscript and to James Farned forthe index.
Don Howard, Roger Hurwitz, Tanya Karachentzeva, Arnon Keren, Last but notleast, I want to thank my family for years of patience and
Alexei Kojevnikov, Edward MacKinnon, Avishai Margalit, Edna support. My husband, Aaron, has alwaystaken a caring interestin this
Ullman-Margalit, Jiirgen Renn, Esther Rosenfeld, Cristoff Schmidt, Zur work, providing uncompromising criticism and technical assistance.
Shalev, Roger Stuewer, and Linda Wessels. My gratitude to Aaron is beyond words.
I gratefully acknowledge the support of the National ScienceFoun-
dation (grant numbers 9011053 and 9123124) and the support of the
 cle vata datgh

CHAPTER 1
ED
. Novelty and Dogma

I believe that to solve any problem that has never been solved
before, you have
to leave the door to the unknown ajar. You have to permit
the possibility
that you do nothaveit exactly right.
Richard Feynman 1998, 26-27

A dialogue can be among any number ofpeople, not just two. Evenone
per-
son canhave a senseof dialogue within himself, if the spirit of the dialoguei
s
present. The picture or image that this derivation suggests is of a stream
of meaningflowing among and through us and betweenus.

David Bohm 1996, 6

Dialogical Creativity

“Science is rooted in conversations.” These words were written

by
Werner Heisenberg, a great physicist of the twentieth century and a
founder of the quantum revolution (1971, vii). How exactly is
science
rooted in conversations? And howdid an extended conversation among
Scientists result in the quantum revolution? These are the major issues
of this book.
4

Science is also rooted in doubt and uncertainty. “Andit is of para-

RES

mount importance, in order to make progress, that we

a Le

recognize this
ignorance andthis doubt. . . . what we call scientific knowledge today
LT LETe A bs He

is a body ofstatements of varying degrees of certainty. Some of them

are most unsure; someof them are nearly sure; but noneis absolute
ly
certain. Scientists are usedto this.” These words belongto another great
scientist of the twentieth century, the quantum physicist Richard Feyn-
man (1998, 3, 27). “We knowthatit is consistent to be able to
live and
not know,” continued Feynman:“I always live without knowing
. That
is easy. How youget to know is what I want to know” (1998, 27-28).
Howis the presence of uncertainty and doubtbuilt into scientific
theorizingat its mostbasic level? How doscientists live without know-
ing? And howdothey get to know? I argue throughoutthis book that
the question of how scientists live in uncertainty, the question of
how
z

2 Chapter One
Novelty and Dogma 3

they create their knowledge, and the question of how science is rooted through which knowledge grows. A dialogical analysis, by closelyfol-
in conversationsare, in fact, one and the same. J elaborate an answer to lowing ideas as they gradually form in numerous dialogues between
this question for the case of the quantum revolution. scientists, deals primarily with the cognitive content of science. It re-
This book examinesthe fluid, open-ended, and often ambiguouspro- quires painstaking attention to every nuance ofthe primary sources. In
cess of scientific creativity, treating it as being rooted in, and perhaps fact, it demands closer attention to the minutia of scientific
indistinguishable from, an ongoingscientific conversation about theo- reasoning
than the older, “internal” (evolutionary) history of science and than
ries, experiments, and instruments. I address the central issue of how “rational reconstruction” accounts.
theoretical knowledge is achieved, articulated, and legitimated. Ialso My exposition differs from the usual accounts by describing the
deal with another major issue: the conceptual and emotional turmoil - _flux of ideas without presupposing underlying conceptual frameworks,
created by attempts to interpret the potent quantum formalism... schemes, or paradigms. In fact, these notions, whether on a global ora
I describe and analyze theintricate flux of dialogues among quan- more restricted scale, are not easily compatible with the dynamics of
tum physicists— dialogues thatresulted in scientific breakthroughsof ceaseless scientific change. Living in doubt and uncertainty is not com-
unprecedented scope andin a radical quantum philosophy. These dia- patible with the accepted historiographical notions of “beliefs” and
logues underlay both the open-minded foundational research and the “commitments.” Nor are Kuhnian and post-Kuhnian “agreement” and
erection of the orthodoxinterpretation of quantum physics: the Copen- “consensus” suitable to describe the dynamicsofliving without know-
hagen interpretation. Tracing the webof dialogues reveals a story about ing. Doubt and uncertainty should be incorporated into the basic terms
the workingsof free scientific imagination and about the consolidation __used to describe the growth of knowledge. From the dialogical perspec-
of scientific dogma. — tive, itis “creative disagreement’—with oneself (doubt) or with others
One of the major puzzles in the history of quantum physics is the (lack of consensus)— thatplays the crucialrole in the advance of know1-
existence of numerous contradictions in the Copenhagen interpreta- edge. The privilege to be unsure,to theorize freely, to explore different
tion. Whatis the sourceof these contradictions? And whyare they im- optionsat the sametime, is incorporatedinto the notion ofcreative dia-
potent to detract.from the spectacular power of quantum physics? A logical flux. I will elaborate on the philosophical, historiographical, and
large part of this bookconstitutes an answerto thesequestions. sociological advantages of such a dialogical approach in the concluding
The analysis and ‘the narrative in this work are permeated with the chapter of the book.
notion of communicability. I argue that dialogue underlies scientific This book simultaneously revises the story of the quantum revolu-
creativity and that the emergence ofscientific novelty cannot be under- tion and outlines a tentative program for a dialogical historiography of
stood without scrutinizing the ways scientists respond to and address science. My work began with a revision of the history of matrix me-
each other. This book analyzes the complex, multidirectional dialogical chanics (Beller 1983) and progressed to revisions of other major epi-
nature of scientific theorizing (part 1) and the strategies by which this sodes in the history of quantum physics, such as the emergence of
dialogical flux is flattened into a monological narrative (part 2). Born’s probabilistic interpretation (Beller 1990) and the birth of Bohr’s
This book is based on a close study of primary sources— correspon- complementarity (Beller 1992a). It gradually becameclear to me that the
dence betweenthe participants, notebooks, original papers. Historians need for ongoing revision has a fundamental historiographical cause.
of quantum physics are fortunate to have access to the Archive for the This cause is intimately connected with the complex dialogica]l nature
History of Quantum Physics (AHQP), where correspondence andorigi- of thought and with the strategies used to flatten it into linear mono-
nal manuscripts are collected. Thus'I had the opportunity to study in logical narratives. :
detail the intricate paths along which ideas emerged as the founders [begin my description of the quantum revolution with an analysis of
of quantum physics addressed each otherin theirletters. I gradually matrix theory in flux (chapter 2), arguing against the received story of
realized that dialogical addressivity permeates not only scientific cor-
theexistence of two totally distinct theoretical frameworks— the matrix
respondencebutalso publishedscientific papers, and more generally, I and the wave theoretical. In chapters 2 and 3 I describe how a strong
came to the realization that scientific creativity is fundamentally dia- distinction between the matrix and wave approaches crystallized as the
logical in the sense elaborated in this book. ; end result of a conceptually fascinating and emotionally intense con-
The dialogical approach tothe history of scienceis “bottom up” —it
frontation among quantum physicists. In the fruitful ambiguity of the
searches for the most basic details in order to conceptualize the process
newly created knowledge, there was no place for strong “beliefs” in
Jesictbeitie

4 Chapter One
Novelty and Dogma 5
indeterminism or “commitments” to positivism. Noris it correct to see The dispensability of paradigmatic interpretive agreem
the matrix theoreticians as committed to a particle ontology, as opposed ent also ex-
plains what Steven Weinberg, not without a touch of disdai
to Schrédinger’s wave approach. Wewill see that Heisenberg’s, Born’s, n, called
“the unreasonable ineffectiveness of philosophy”in scientific
and Pauli’s pronouncements on foundational and interpretive issues practice
(Wginberg 1992, 169). The application of theoretical tools involve
wereall fluid and uncommitted. Similarly, I will argue, Born’s prob- sflexi-
bility and conceptual opportunism—no system of philosophica
abilistic interpretation did not stem from his “belief” in particles and l pre-
conceptions can survive in such a fluid environment (chapter
“commitment” to indeterminism, as the received history of quantum. 3). Philo-
sophical “influence” indeed cannot determine a scientific
physics implies (chapter2). knowledge
claim. Yet philosophy can be suggestive in limited way, as inspir
The flux of ideas in the emergence of matrix theory andin the for- ation
along some path in the dialogical web of creativity. In chapter 4 ar-
mation of Born’s probabilistic interpretation demonstrates the primacy gue that the ideas of the German philosopher Fichte were importantfor
of mathematical tools over fundamental interpretive ideas. Thesetools Heisenberg’s treatment of measurementin the uncertainty paper.
-can be borrowed, developed, and successfully applied withouta clear- This
treatment, or “reduction of wave packets,” became the source
cut stand on basicinterpretive issues. It was on the efficiency of the of the
major conceptual hurdle of quantum physics—the notorious measure-
mathematical tools, and not on metaphysical “paradigmatic” issues, ment problem.
that there was agreementin the community of quantum physicists. And Scientific creativity as a dialogical flux is exemplified by the emer-
on this point the orthodox and the opposition were united; agreement gence of Heisenberg’s uncertainty principle, which I describe in chap-
on the potencyof these tools prevented scientific practice from disinte- ter 4. We can see how Heisenberg theorized without a clearly delin-
_ grating, be the philosophical disagreements as large as they may. eated conceptual framework, without “beliefs” and “commitments.”
Theoretical tools (equations, methods of solution, and approxima- Instead, we observe the indispensability of open-ended disagreement
tions) have their own momentum,while philosophical ideas are adap- and ongoing doubtfor achieving a theoretical breakthrough. The ex-
ted a posteriori (chapter 3). The fact that theoretical tools have some isting historiography presents one-dimensional pictures: Jammer de-
autonomy allowsscientists to theorize without taking an interpretive _Scribed the uncertainty formula as emerging naturally from the impera-
stand. The English physicist Charles Darwin wrote to Niels Bohr: “It is tive to adapt the new mathematical formalism’ to thé possibilities of
a part of my doctrinethat the details of a physicist’s philosophy do not measurement (Jammer 1966).* The perspective offered by social histo-
matter much” (Darwin to Bohr, December 1926, AHOP). This belief in tians similarly focuses only on single aspect: a predilection for acau-
the primacy of mathematical tools was especially strong in Gottingen, sality among quantum physicists determined their efforts to interpret
inspiring Born’s and Heisenberg’s search for a new quantum theory.

csereaeie
the new quantum mechanics along probabilistic lines (Forman 1971;
“Mathematics knowsbetter than ourintuition” was Born‘s motto (inter- Feuer 1963).
view with Born, AHQP). I describe Heisenberg’s discovery of the uncertainty relations as a
The neutrality of a mathematical formalism with respectto possible multidirectional process, which took place in a communicative network
interpretations has another far-reaching consequence: scientists may with many interlocutors, including such prominent names as Einstein,
give all authority in interpretive matters to a few leaders, whose phi- Schrédinger, Pauli, Dirac, Bohr, Born, and Jordan. Less known, yet
losophytheyare willing to accept. Such humble resignation from philo-
h

no less important for Heisenberg’s discovery, are the names Campbell,

a

sophical exploration is often nothingbut a convenient choicenot to deal Duane, Zernike, and Sentfleben. In chapter 4 we follow the process
“with confusing and perhapsirrelevant matters.It is this attitude that of discovery and observe how fragments of insight gradually emerge,
creates room for an authoritative and privileged interpretation, such as how ideasclash, change, disappear, or survive. The gradual forming
the Copenhagen orthodoxy. of
a preference, of choosingoneintellectual option over another, of defin-
ing whatthe options are—all occurin a coalescence of insights, arrived
1. This position was taken by Darwin: “I’m quite ready in advanceto believe that your at in different dialogues andat differenttimes.
criticisms are quite right, but I feel that perhaps this does not matter much. Because the
best sort of contribution that people like me can make to the subject is working out of Heisenberg’s discovery of the uncertainty principle was a complex
problems, leaving the questionsofprinciples to you. In fact even if the ideas on whichthe
2. Jammer described the controversy between Bohr and Heisenberg with respect to the
work was done are wrongfrom the beginningto end,it is hardly possible that the work
uncertainty paper as a toolfor elucidating Bohr’s and Heisenberg’s respective
itself is wrong in that it can easily be taken over by any revised fundamental ideas that positions
and as anecdotalhistory. From Jammer’s analysisit is not clear how the disagreeme
you maymake” (Darwin to Bohr, December 1926, AHOP). nts
contributed to the creationof uncertainty and complementarity.
rtetnd

6 Chapter One Novelty and Dogma 7

process of disagreements, qualifications, elaborations, supplementa- Rimini, and Weber 1986). The flux of creative research cannct be forced
tions, and borrowings. He had no foundational commitments, even on into an unequivocal, final conceptual scheme.
such basic issues as discontinuity and indeterminism; rather his pref- The description of Heisenberg’s creative theorizing calls for a re-
erence for discontinuity and acausality took shape gradually in many evaluation of the role of “lesser” scientists in the growth of scientific
fruitful dialogues. Though immersedininterpretive efforts, Heisenberg knowledge (chapter 4). We will see that some of the most important
was uncertain even about whatthe word “interpretation” means(chap- insights pertaining to Heisenberg’s formulationofthe uncertainty prin-
ter 5). Disagreements with interlocutors—a militant one with Schré- ciple belonged to scientists whose names do not appearin the received
dinger, a subdued yet painful one with Bohr, a restrained one with Jor- history of quantum mechanics—Sentfleben and Campbell. The issueis
dan and Born were triggers for Heisenberg’s reasoning. Agreement. not one ofpriority, even thoughthedistributionofcredit is often unfair.
too played a part, in the form of Heisenberg’s partial, often only tem- ~ ' Neither Campbell nor Sentfleben, from their positions in the commu-
porary, acceptance of the ideas of others, especially those’ of Dirac, nicative web, could have accomplished exactly what Heisenberg did.
Campbell, and Duane. Heisenberg’s case demonstrates how genuine Yet neither could Heisenberg have developed his ideas had he not been
novelty emerges through dialogical creativity. Dialogical creativity is respondingcreatively to Campbell’s and Sentfleben’s insights. Heisen-
not an instantaneous “eureka” experience;it is rather a patiently sus- berg’s discovery is organically linked to the ideasof the “lesser” scien-
tained process of responsiveness and addressivity to iis ideas of others, tists. From the epistemological point of view, the notion of a scientific
both actual and imagined. collective is intrinsic to the dialogical approach.
One might expect that in a published scientific paper, all previous *_.We can reevaluate the prevalent ideaof a lonely creative individual,
cognitive tensions would beresolved and a coherent unequivocal mes- _and of solitude as a precondition of creativity. Conventional opinion
sage expressed. Yet an analysis of Heisenberg’s uncertaintypaper finds holds that “the spark ofcreativity burns mostbrightly in a mind work-
clear traces of past struggles, conflicting voices on the same issue, and ing in solitude” (Storr 1988). When Heisenberg “fabricated” the new
unresolved tensions (chapter 5). The polyphonyofthe creative act ech- quantum mechanics(his expression, van der Waerden 1967, 15), he was
oes in the’ paperitself. Similarly, at least twoconflicting, in fact incom- isolated on the island of Helgoland. After their stormy debates on the
patible, voices can be heard in Bohr’s response to Einstein-Podolsky- interpretation of quantum mechanicsin the fateful year of 1926, Heisen-
Rosen’s challenge to the Copenhageninterpretation(chapter 7). berg and Bohr needed to get away from each other. Separated, Heisen-
One might object, then, that suchunresolved tensions perhaps char- berg wrote the essentials of his uncertainty paper and Bohr elaborated
acterize scientific papers written during revolutionary upheavals, but his complementarity. Yet solitude does not imply cognitive isolation.
whenthings settle down and the revolution is over, a new paradigm _If Heisenberg needed time away from Bohr,it wasin orderto strike a
triumphs and the foundational debate is closed. There is, the argu- proper, uncoerced balance in his own communicative network of cog-
ment might continue, at the present time onlyone correct, agreed-upon nitive responses (chapters 4 and 6).
meaning of the uncertainty principle and of wave-particle complemen- The dialogical perspective provides a new wayto read publishedsci-
tarity. But is there? One can find a “correct’” meaning in textbooks, or in entific texts. Concealed doubt becomesvisible, and a paper becomes a
some philosophical writings cn the quantum theory—in short, in the fascinating scientific and human document, resounding with conflict-
graveyards of science. On the research frontier nothing is immune to ing inner voices, populated by many“virtual” interlocutors (chapters 5
reappraisal—be it uncertainty, complementarity, or even the determin- and 6). We will see in chapter 5 that Heisenberg’s uncertainty paperis
ism of classical physics (chaos theories) and the indeterminism of quan- permeated with doubts and indecision on such central issues as inde-
tum physics (Bohm’s theory and Bohmianalternatives, such as Diirr, terminism, realism, visualizability, and the status of classical concepts
Goldstein, and Zanghi 1992a, 1992b, 1996). Numerous meaningsof the in the quantum domain. A comparison of the (seemingly) confident
uncertainty formulas are proposed in current research papers (Home published paper with the almost identical draft (in a letter to Pauli)
eemeeCe

and Whitaker 1992; Valentini 1996). The sameis true of Born’s probabi- filled with doubt regardingall the basic interpretive issues reveals how
listic interpretation andhis solution of the collision problem (Daumer misleading it is to ascribe any “beliefs” or “commitments” to Heisen-
1996). Similarly, there is no agreement about the meaning or even the berg. My analysis applies to the issues of acausality and positivism that
validity of wave-particle complementarity (chapter 11). Even such basic according to the accepted history of quantum physics, are the two cen-
formulas as Schrédinger’s equation are opento modification (Ghirardi, tral pillars of Heisenberg’s uncertainty paper.
 ia iatali
Mees |

8 Chapter One
Novelty and Dogma 9
My examination of Heisenberg’s uncertainty paper reveals the
argu- mechanics). These first attemptsat interpretation revolved around
mentative strategies by which interpretive freedom is conceal piv-
ed, and otal experiments by Franck and Hertz (1913), Bothe and
the illusion created that the orthodox interpretation is “inevitable” Geiger (1925a,
— 1925b), and Compton (1923), as well as around experim
the issue to which the bulk of part 2 of this book is devoted. Tension ental work by
between the conceptualfreedom experienced by the Copenhagen e lesser known Moll and Burger (1925). There is a crucial
phrs- difference
icists and their desire to advocate one privileged interpretation between evidence-based interpretive efforts and closed
is one dogmatic sys-
of the major sources,I argue, of the numerous contradictions tems, althoughthe first can degenerate into the second,
and incon- as with Bohr’s,
sistencies in the Copenhageninterpretation of quantum Heisenberg’s, and Pauli’s later philosophical writings, which
physics. cameclose
Yet the dialogical analysis of a published paper does more than to preaching a rigid ideology (part 2).
reveal Chapter 7 (the last chapter of part 1) analyzes the process by which
the inconclusive nature of scientific theorizing. Such an analysis can
also substantially modify our understanding of the content of a pub- sincere and open-minded, though interest-laden, interpretive attempt
s
lished paper. A dialogical readingis a potent tool for deciphering hardened into an ideological stand intended to protect quantu
es- m theory
pecially obscure and opaquetexts. Chapter 6 is devoted to a dialogi from challenge andcriticism. This chapter, partly based on my
cal paper
analysis of Bohr’s Comolecture, in which Bohr announcedhis with Arthur Fine (Beller and Fine 1994), is devoted to an
principle analysis of
Bohr’s reply to the Einstein-Podolsky-Rosen (EPR)challenge. This chap-
of complementarity.
- The Como lecture is considered one of the most incomprehens
ter is located, so to speak, ona “cut” between parts 1 and 2.and can be
ible moved from the end ofpart 1 to the beginning of part 2 without dis-
texts in twentieth-century physics. My dialogical reading constitutes a
basic revision of the accepted reading of this text, by presenting the torting the argumentof my book. This means that dialogical emergence
Comolecture, not as the unfolding of a single argumentative struc- and rhetorical] consolidation are not completely distinct processes. Con-
ture, but as the juxtaposition of several simultaneously coexisting argu- solidation was already present in the initial interpretive attempts of
ments, addressed to different quantum theorists aboutdifferent issues. the Géttingen-Copenhagen camp, andlimited dialogical responsive-
A dialogical analysis reveals that the central message of Bohr’s paper ness accompanied later elaborations of the Copenhageninterpretation.
Still, chapter 7 reveals a vast difference between Bohr’s reasoni
wasnot the resolution of wave-particle duality by the complementarity ngin his
principle, as usually assumed,but rather Bohr’s extensive defense ofhis Comolecture and his reply to the EPR challenge. Myanalysis uncovers
concepts of stationary states and discontinuous energy changes (quan- a transition from legitimate, though often confused, argumentsfor the
tum jumps) against Schrédinger’s competitive endeavors. consistency of quantum theory, to argumentative strategies promot
ing
A dialogical reading allows the reevaluation of some central the inevitability of the orthodox stand. This transition naturally con-
events tains both old insights and new conquests, andit is not surprising that
in the history of quantum physics, such as the famous clash betwee
n two different, often incompatible voices permeate Bohr’s response to
Heisenberg and Bohr over the uncertainty paper. This reevalu
ation EPR. The voices correspond to Bohr’s two different answers to EPR: one
merges “motives” and “reasons,” connecting “conceptual” and
“anec- relying on the concept of disturbance,the other dispensing with it. In
dotal” history into one meaningful, comprehensible story. This
story contrast to the received story, which, following the orthodox narrative,
throws new light on Einstein’s and Schrédinger’s initial reservations
about the early interpretive attempts of Bohr and Heisenberg affirms Bohr’s “victory”in this confrontation, the analysis in chapter 7
—reser- reveals that none of Bohr’s answerscan be considered satisfactory.
PF UUSEEPROELONS IPTG NOMINEES a “ae
vations due to difficulties and contradictions in the emerging Copen-
hagen interpretation rather than to Einstein’s and Schrédinger’s Ifurther analyze Bohr’s about-face on the central interpretive issues—
“con- the problems of reality, acausality, and measurement—raised by
servatism.” the
Myanalysis of the initial efforts of Bohr and Heisenberg EPR challenge. I extendthis analysis in chapter8, arguing that the Co-
to unveil penhagen interpretationis in facta compilation of various philosophical
the physical meaning of the quantum formalism demonstrates
the vast strands, given a public presentation that often hid shifting disagree-
freedom of interpretive endeavors.Yet this freedom is
not arbitrar y. We ments betweenits main architects. The Copenhagen interpretation of
will see the great extent to which the formulations of Born’s
probabi- quantum physics did notoriginate from a disinterested search for philo-
listic interpretation, Heisenberg’s uncertainty, and
Bohr’s complemen- sophical foundations—from the very beginning it was constructed in
tarity were woven arounddetailed analyses of experimental
situations the heat of a fierce confrontation. As the nature of the opposition’s chal-
(a fact not sufficiently apparent in the existing histories
of quantum lenge changed, so did the local responses of the orthodox. It is not
 TETTT NEEN
a es

10 Chapter One Novelty and Dogma 11

‘ete aN

surprising therefore that whatis called the Copenhagen interpretation opposition and the description of past science concurrently. The ideas
is so riddled with vacillations, about-faces, and inconsistencies (chapter

SMILENTT
of the opposition are projected as most characteristic of the overthrown
8). The orthodox aimed to present a united frontto the opposition, con- past, and thus the opposition naturally appears reactionary—dispos-
cealing the substantial differences of approach among its members. I ing of the old and discrediting the opposition are, in fact, one and the
analyze some of the strategies by which such differences were sup- same process. As a result, not only is the opposition caricatured but
pressed byrelying on a distinction between whatscientists “must not” past science is trivialized. Hero worship of the winners further dele-
and whatscientists “need not” do (chapter8). gitimates the opposition and preventscriticism of the orthodox stand
‘(chapter13).
Rhetorical Strategies we Historians of science rarely question the narrative of the winners:
Jammer(1966) and Mehra and Rechenberg (1982), for example, closely
Howdoes one construct from among these numerouscontradictory ar- follow the orthodox line. Jammer, in the preface to his classic book,
guments a narrative that seemsto irrevocably implythepillars of the quotes Einstein’s penetrating warning not to rely on the recollections
Copenhagen dogma? How does onereconstruct history so that the of the participants: “To the discovererin this field the products of his
central tenets of the Copenhageninterpretation, such as indetermin- imagination appear so necessary and natural that he regards them. . .
ism and the impossibility of an objective, observer-independent de- not as creations of thought butas givenrealities.” Yet Jammer“felt en-
‘scription, seem not merely highly persuasivebut outrightinevitable? titled to ignore this warning,” and he “discussed the subject with quite
In part 2 of the book I contend thatall the Copenhagen arguments of a number of prominent physicists who contributed. decisively to the
“inevitability” are in fact fallacious—they rely either on circular rea- developmentof the theory”(1966,viii). Pais’s recent book (1991) is writ-
soning or on highly appealing but misleading metaphorical imagery ' ten exclusively from Bohr’s perspective. This is not to say that these ac-
(chapters 9 and 12). They are strongly supportedbyfalsified, history, counts do not use primary sources; in fact, Jammer, Mehra and Rechen-
which renders certain developments as dictated by the inner logic of berg, and Pais use the sources extensively. Yet what they see in those
the development of ideas (chapters 10 and 11). Discrediting the oppo- sources, and more important, what they ignore therein, is dictated by
sition and caricaturing the opposition’s criticism of the Copenhagen the overwhelming authority of the winner’s perspective. A notable ex-
‘stand is’ yet another- potent rhetorical device to strengthen the ortho- ception in this respect is Cushing’s (1994b) important book about Boh-
doxy (chapter 13). These chapters reveal how fruitfully ambiguous and mian mechanics as a viable alternative to the Copenhagen hegemony.
wisely uncommitted interpretive efforts are concealedbyrigid recon- The “naturalness” and even “finality” of the orthodox point of view
structed stories. Complex, many-voiced, multidirectional theorizing is is advanced through powerful strategies of persuasion, which refer to
thus conflated into an orthodox, one-dimensional narrative. as the “rhetoric of inevitability” (chapter 9). The ingenious technique
“History is written by winners’”—this cliché finds powerful confir- wasto disguise arguments of consistency as thoseof inevitability. What
mation in the case of the quantum revolution. We have numerous remi- is taken as objective quantum philosophy (and “inevitable” at that)
niscences by the winners—Bohr, Heisenberg, Born, Jordan, and others. turns out to be ideology—where by ideology J mean a system of as-
There are hardly any reminiscences by Einstein and Schrédinger about sertions that imply, from within, their own justice, truth, and self-
the same events—wedo nothear the opposition’s side ofthe story. In evidence. With respect to the entrenchmentof the Copenhagen dogma,
the quantum revolution, the orthodox constructed the narrative, elimi- epistemology and sociology often merge—considerations of epistemic
nating dissident voices and largely suppressing the crucial contribu- warrant and of social legitimacy are, at times, indistinguishable. The
tions of the opposition andoflesser scientists. In part 2 I describe the foundations of the Copenhagen paradigm werechosen and elaborated
4
=
strategies by which the past is manipulated in order to make the win- in direct contrast to the opposition’s stand. The construction of the win-
2
co
ners look naturally right. By such a reconstruction of the past, the
ner’s narrative and philosophical arguments of inevitability serve the
cornerstones of the Copenhagen interpretation—quantum jumps, the same end.
impossibility of causal space-time models, indeterminism, and wave- In chapter 11, I contrast the dramatic narrative of the “inevitability”
= particle complementarity—were even morefirmly entrenched (chap-
of wave-particle complementarity with the freedom and plurality
ters 10 and 11).
of theoretical approaches to the wave-particle issue. The “logical”
I describe howthe opposition’s standis delegitimated andtrivialized. arguments for the inevitability of wave-particle complementarity are
In their fabricated narratives, the winners construct the profile of the
built on Bohr’s peculiar doctrine of the “indispensability” of classical
 12. Chapter One
Novelty and Dogma 13
concepts—a doctrine that few theorists, including Bohr’s
closest colla- guided the search for the new quantum theoryin the early 1920s, and
borators, found persuasive (chapter 8). Those mathematical physici with which the complementarity principle is sometimes confused. The
sts
whogaveupthis rigid doctrine suggested a rich variety of solutio
ns to metaphors of the complementarity principle are vague andarbit
rary, in
the wave-particle dilemma (chapter 11). The discussion in chapte
r 11 contrast to the more rigoroususe of analogies between the macro- and
demonstrates that there is a fundamental ambiguity, and therefo microdomains guided bythe correspondence principle. While the cor-
re a
lack of “paradigmatic” agreement, concerning even such
basic physical respondence principle was a potentheuristic that led to the discove
terms as “wave” and “particle” (scientists hold different opinion ry
s, each of the rigorous quantum formalism,the complementarity principle
was
from his own theoretical viewpoint, of what are to be consid a device of legitimation— it led to no new physical knowledge.
ered nec-
essary or sufficientattributes of those terms). Despite this lack of clarity, In chapter 12, I analyze the ways Bohr, Pauli, and Heisenberg, by
or perhaps becauseof it, conversation onthis issue continued, their imprecise allusions to quantum “wholeness,” spun a metaphorical
and new
theoretical breakthroughs occurred. webof associations that disguise, rather than reveal, quantum entangle-
The same freedom that created room for open-ended creative theo- ment and nonlocality. I argue that these allusions and analogies are fed
tizing allowed the construction of a variety of ad hoc strateg
ies for by classical intuitions and contain nothing quantum about them. Not
the legitimation ofthe Copenhagenstand. Theproliferation of such ad surprisingly, Bohm’s version of quantum theoryandits recent variants,
hoc movesis yet another source of the inconsistenciesthat still plague which fundamentally incorporate quantum wholenessas a basic prin-
the Copenhageninterpretation today. Both Heisenberg and Pauli sup- ciple, are compatible with deterministic description. Bohr’s numerous
ported Bohr’s philosophy of wave-particle complementarity in public, opaqueallusions to quantum wholenesscontribute to the illusion that
while often expressing, behind closed doors, views that were contrary his philosophical views were stable, despite the fact that he used this
to Bohr’s. Heisenberg, Born, and Pauli, as well as Bohr notion differently in different contexts. The notion of wholeness un-
himself, ex-
ploited wave-particle complementarity for pedagogical and ideologi- dergoes especially drastic change between Bohr's pre-1935 and his
cal reasons. The simple thought experiments that supposedly demon- post-1935 writings (as aresult of EPR). This ingenious and misleading
strated the necessity of both the wave and particle descriptions were improvisation on the idea of quantum holism contributes to the decep-
especially effective for promoting the philosophical “lessons” of quan- tion that a well-defined philosophical framework exists, thus further
tum theory to wider audiences. From the analysis.of these experiments, entrenching the Copenhagen orthodoxy.
supported by Bohr’s doctrine of the indispensability of classical con- Support for the orthodox view comes not only from historians but
~
cepts, both the “finality” of indeterminism and the “impossibility” from philosophersof science as well. The strongest support came from
of
unified objective description followed, leaving seemingly no possibil Norwood Hanson and, after him, Thomas Kuhn, who incorporated
ity
of other interpretive options. the Copenhagen ideology into an overarching theory of the growth of
The accessibility of the explanatory strategies fed the illusion scientific knowledge (chapter 14). Hanson and Kuhn canonized the
that
no technical understanding of the quantum mechanical formalism concepts of paradigm and incommensurability intoobjectified philo-
is
needed in order to grasp the essenceofthe quantum revolution. sophical notions that exclude,in principle, diversity of opinion andle-

Ser ET TTD RCO

This
illusion was most vigorously sustained by Bohr himself. “Philo gitimate disagreement. Thus oppositionis discredited and eliminated
sophy
is but a sophisticated poetry”—this view of philosophy aptly in the most radical way—by definition. Kuhn and sociologists of sci-
char-
acterizes Bohr’s voluminous improvisations on the theme ence who follow Kuhn’s approach talk in terms of “deviance” and
of. comple-
mentarity, filled with affective analogies, subjective associations, “impermissible aberration” rather than acknowledging reasonable dis-
and
allusions to “harmonies,” expressed in “common languag
stein bate connec onan RATE

e.” In chap- agreement. Kuhn’s theory of incommensurable paradigms and the or-
ter 12, I argue that Bohr’s philosophy is best characterized thodox narrative of the quantum revolution reinforce each other—the
as a richly
imaginative, yet ultimately misleading attempt to compre quantum revolution is cited as a prime example supporting Kuhnian
hend the
quantum mystery without recourse to the mathematical formal philosophy,* and the orthodoxnarrative of the quantum revolution is
ism of
quantum theory. Thereis, in this sense, a vast
difference between the
complementarityprinciple and Bohr’s correspondence princip 4. Even historians who question the adequacy of Kuhnian notions to describe other
le, which
historical developments (Westman 1994, discussing the Copernican-Newtonian revolu-
3. Many postmodernistcritics of science fell prey to the temptati tion) assume that Kuhn’s philosophy adequately describes the case of the quantum revo-
ons of this strategy of
argumentation (Beller 1998).
lution.
 LLLPTEET eT
14 Chapier One

objectified by a Kuhnian theory of the growth of knowledge (chap- PART ONE

ter 14). By such circular argumentation, the orthodox perspective is oF
madeto appear unassailable.
By incorporating addressivity and disagreementas fundamental no-
tions, the dialogical approach, developed here, presents an alternative

ae
to current approachesto the studyof science. Dialogical analysis incor-
porates conversation and communicability both as social realities and
as epistemological presuppositions. From the dialogical perspective, a
creative scientist cannot, in principle, be isolated—he,or she,is linked
fundamentally to the efforts and concerns of others. In the dialogical
approach, the notorious question of whether science is “tational” or
“social” in nature becomes a pseudoproblem.Scienceis simultaneously
rational and social—the rationality of science is dialogical and com-
municative. Dialogical Emergence
The viewof scientific activity as an snlocteapdies open-ended com-
municative flux fits well scientists’ own image of their work. Theoreti-
cal physicist David Finkelstein (1987), describing the state of his dis-
cipline today, chose the Heraclitean “All is flux” for the title of his
paper. David Bohm emphasized the essential communicability of sci-
entific doing: “Communication playsan essential role in the very act of
scientific perception. . . They [scientists] constantly engage in a form
of internal dialogue with the whole structure of their particular disci-
pline. . .. When insight occurs, it emerges out of this overall structure
- of communication and must then be unfolded so thatit obtains its full
meaning within it’ (Bohm and Peat 1987, 67).
In the concluding chapter of this book I providea tentative outline for
a dialogical historiography and philosophyofscience. Dialogical histo-
riography reestablishes scientific individuality as the focus for studies
of the growth of knowledge. Dialogical analysis demonstratesthat sci-
entific theorizing can be both free and nonarbitrary, andthattheoretical
achievements can be earsReceTeAeTEEETG well grounded and imaginatively
beautiful.
"

We constantly conduct conversations with others—with living

pinailieabedaisdesBaldi. cta dh
oe

people, with the dead, and even with the yet unborn. The Russian poet
Marina Tsvetaeva, in one.of her poems, addressed a reader one hun-
dred years in the future—the one whowill truly love and understand
her, and whowill prefer her remainsto the flesh ofthe living. Without
unceasing addressivity and communicability, existence and thought,
artistic imagination andscientific creativity are inconceivable.
 CHAPTER 2
ea

Matrix Theory in Flux

Antagonistic cooperation is the key.

William Carlos Williams 1936, 177

Introduction

The years 1925-27 were a time of astoundingscientific creativity. Dirac

(1977) referred to this time as the “exciting era,” while Pauli cailed
it a
“period of spiritual and human confusion” (1955a,30). It was both. The
“exciting era” refers to the breathtaking advances in the foundations of
physics, “human confusion”to the treacherous maitter of interpretation,
to intensely emotionalconfrontations, to the slippery philosophical im-
plications of the new conceptualtools. In this period of unprecedented
advance the openness and ambiguity of theoretical practice seemed
“confusing.” But perhaps it was this ambiguity that created conceptual
room for such rapid progress. There was no strong “belief” either in
indeterminism or positivism, as the received story implies—pronou
nce-
ments on these issues were uncomunitted, fluid, and’ opportunistic.
The human confusion is reflected in the rich folklore of the history of
quantum physics— anecdotes about inflamed passions and wounde
d
egos, about emotional strain and outspoken hostility. These emotions
produced not merely heat—they begotlight (Beller 1996b).
:£ In this chapter I analyzé theinitial efforts to interpret the new ma-
& trix formalism.I also describe the emotional confrontation between the
&
é
matrix physicists and Schrédinger, which fused, or confused, issues
of
foundational preference with that of professional privilege. largue that
the emergence of Born’sstatistical interpretation of quantum
arora Pe RSEEIEN SAT RT IE RT

= mechan-
icswas devoid of prior philosophical preferences and was embedded
RC its

in the dialogical context of the “battle” between Schrédinger and the

Gottingen-Copenhagen camp.
This chapter and the following one undermine the myth that the
matrix and wave approaches were from the beginning conceptu
ally
distinct and historically independent. I argue that the strong distinc-~
tion, and even polarization, of the wave theoretical and matrix
approaches developed as a result of the confrontation between the
 18 Chapter Two Matrix Theory in Flux 19

competing groups.' This being the case, physicists on both sides were Contrary to this story, | have argued that matrix mechanics was not

YUE: Betora gDNEet

not generous in their papers and later recollections with acknowledg- immediately recognized as a pivotal breakthrough(Beller 1983). A few
ments of mutualdebts. months before the publication of Schrédinger’s mechanics andits use
in transformation theory, matrix mechanics wasbesetbydifficulties of

ihe:
A Revision of the Origins of the Matrix-Theory such magnitude that no one,including the authors themselves, consid-
ered it to be morethana first step on a long path towardtheultimately
Thestill widely accepted scenario for the development of the new correct theory. This was the main reason for the enthusiastic acceptance
quantum mechanics in the 1920s runsasfollows. In a period ofonly a of Schrédinger’s theory: acclaim for Schrédinger’s theory wasnotlim-
few months(late 1925 to early 1926), two majortheories of atomic phe- ited to the conservative quarters of the physical community... |.’
nomena emerged: matrix mechanics and wave mechanics. Matrix me- Nor was matrix mechanics a theory of corpuscles before Born’s prob-
chanics originated in Werner Heisenberg’s radical reinterpretation of abilistic interpretation: an atom in the matrix approach was endowed
basic physical magnitudes (Heisenberg 1925), and Max Born, Pascual with electromagnetic, not kinematic, meaning. It is impossible to un-
Jordan, and Heisenberg immediately expandedit into a complete and derstand the genesis of the philosophy of quantum mechanics with-
logically consistent theoretical structure in the so-called Dreimanner- out closely following the original interpretive announcementsof matrix
arbeit (Born and Jordan 1925a; Born, Heisenberg, and Jordan 1926). physicists and the modifications they later underwentin dialogical re-
Wolfgang Pauli (1926b) demonstrated that the abstract formalism of sponse to Schrédinger’s theory. The radical assumption of the matrix
matrix mechanics gave the correct approach to atomic theory bysolving approach—within the atom there is no- geometry—evolved into the
the problem ofthe structure of the hydrogen atom—thecentral prob- more moderate: within an atom there is only statistics. The classical
lem of the atomic domain. The basic conceptof matrix theory was the space-time container, eliminated by the matrix approach, was restored
particle, or corpuscle.? | eo. in response to Schrédinger—first by Born, then by Pauli and Jordan,
The basic concept of the rival theory, Erwin Schrédinger’s (1926a, and finally by Heisenberg himself. It is easy-to overlook this process
1926b, 1926c, 1926d, 1926e, 1926f), was instead the wave, based on ideas because of the speed with which it took place. The resulting quantum,
first proposed by Louis de Broglie (1923). Schrédinger’s theory com- philosophy evolved into a hybrid of the original radical matrix pro-
bined intuitive and adequately developed mathematical tools with fa- gram, concepts revived from Bohr's earlier work, andstatistical com-
miliar physical concepts, and it was enthusiastically welcomed bythe promises necessitated by the acceptance of Schrédinger’s continuous
conservative wing of the physical community, which distrusted the theory. The philosophy of quantum physics became a mixture of quan-
revolutionary physical ideas of matrix mechanics and the complicated tum and classical concepts, whose inherent difficulties plague physi-
mathematics involved. Thoughradically differentin their basic assump- cists and philosophers to the present day.
tions and mathematical treatments, the two theories were soon proved
to be equivalent by Schrédinger himself? and the only major problem Problems of Physical Interpretation: The Elimination of Space-Time
left was that of physical interpretation. Schrédinger attempted to as- Matrix mechanics, designed to avoid the problemsof the old quantum
cribe to his theory a classical, continuous wave interpretation, but the theory, implied a radical change in the converitional description of the
quantum physicists at Géttingen and Copenhagen demonstrated that

serenaseentronenntneei e PT CHa reeSPREE

space-time continuum used in earlier physical theories. Heisenberg’s
these conservative attempts were untenable. The physical interpreta- 1925 paper—the turning point in quantum physics—announced the
tion that prevailed instead was a direct continuation of the indetermin-
| = xxl
istic beliefs of Heisenberg, Born, and Niels Bohr.
aim of eliminating unobservables and dispensing with visualizable
= modelsthat relied on continuous space-time pictures. Instead of such
| z
=
1. This book analyzes the Géttingen-Copenhagenside of the confrontation. An analo-
unobservable kinematic variables as an electron’s position, velocity, and
period of revolution, Heisenberg sought to incorporate only observable
gous argument can be made about Schrédinger.It is explored in part in Wessels (1983). I
4
am currently exploring Schrédinger’s case as well. spectroscopic data into the theoretical framework.
*,
2. See also Darrigol (1992a), Jammer (1966), and Mehra and Rechenberg (1982). Heisenberg andhis colleagues at Géttingenjustified their approach
= 3. Recently, Muller (1997) challenged the “myth of equivalence.” Heclaimedthatthere
| 4@ was no strict equivalence between Schrédinger’s and Heisenberg’s versions originally— 4. Although Heisenberg did not rely on it explicitly, the “virtual oscillator” model,
the two theories becametruly equivalent after von Neumann’s formalizationoftransfor- whichcontradicted the usualpicture of space-time within the atom, played an important
mation theory.
: heuristic role (MacKinnon 1977).
20 Chapter Two
Matrix Theory in Flux 21

by arguing that the difficulties of the old quantum theory stemmed

a change ofposition in time butalsothat it is impossible to ascribe po-
from the fact that the quantum calculation rules operated with unob-
sition to an electron at a given instant. Space-time exists only in the
servable quantities—they were based on impermissible classical me- macroscopic domain—in the atomic domain “spacepoints in the ordi-
chanical pictures, the use of which madetheold theory inadequatefor nary sense do notexist’” (Born 1926a, 128).
solving any but the simplest cases andled to grave innercontradictions Pais approach was tantamountto giving up all hopeof devising a
(Heisenberg 1925). The Géttingen program wasthen to replace the ex- visualizable physical interpretation for the new matrix mechanics. As
planatory mode of continuous classical mechanics with a discrete de- Jordan putit later, there is no reason to expect that mathematical rela-
scriptive approach. i, tions between observable radiation entities are amenable to visualiza-
Born had madea pivotal contribution toward a “truly discontinuous” tion built on classical habits (1927a, 648). Matrix mechanics is no more
theory byinventing a method(used in Heisenberg’s reinterpretation pa- opento visualization than Maxwell's equations —one can onlyget used
per)of replacingall differential coefficients with the corresponding dif- to manipulating both. The only physical interpretation to be expected.
ference quotients (Born 1924; Born, Heisenberg,.and Jordan 1926). This from such an approachis the description of the limiting case of transi-
replacement reflected the growingrealization that in the new quantum tion from theessentially unvisualizable quantum domainto the visual-
theory each physical quantity should depend on twodiscrete stationary izable macroscopic world."
states, and not on one continuousorbit as in classical mechanics.5 The Even though Heisenberg andhis colleaguescould dispense withthe
old quantum theory chose certain stationary states from all possible demand for visualization on the atomic level, they still needed ta com-
mechanical motions by meansof quantization rules. The new theory, prehendthe transition from microdomainto classical macrodomain.
Born emphasized, was to contain only the permissible values: consis- Yet the problem proved unyielding, and Heisenberg complained to
tency demandedthat forbiddenfractional values should have no mean- Pauli abouthis failure to understand how to achievethetransition.”
ing at all in the theory. ne One ofthe principal reasons for this difficulty was the purely formal
This approach implied the elimination of classical space-time as a definition ofthe concepts of time and space in the matrix theory, which
container of motion—sucha containeris as superfluousas the unreal- bore norelation to evolution over time in a physical system. Pauli had
ized mechanical motions: “No onehas beenable to give a method for mentionedthis fact as early as November 1925 in a letter to Bohr. Cor-
the determination of the period of anelectronin its orbit or even the nelius Lanczoslater argued that an atom in the matrix approach is a
position of the electron at a given instant. There seemsto be no hope timeless entity (1926, 815). Heisenberg agreed that the concept of time
that this will ever becomepossible, for in order to determine lengths or has no more the usual meaning in the new theory than the concept of
times, measuring rods and clocks are required. The latter, however, space—both are merely formal symbolic artifacts.
consist themselves of atoms and therefore break down in the realm of
atomic dimensions” (Born 1926a, 69). 6. “In the further development of the theory, an importanttask will lie... in the man-
Heisenberg admitted later that he intended to eliminate not only the ner in which symbolic quantum geometry goes overintovisualizableclassical geometry”
orbits of bound electrons but even the experimentally observable paths (Born, Heisenberg, and Jordan 1926, 322). :

aa
7. “The worstis that it does not become clear to me how the transition into classical
of free electrons, despite Wilson’s experimental substantiation of the
theory takes place” (Heisenberg to Pauli, 23 October 1925, PC, 251).
latter (Heisenberg 1933, 292). The authors of matrix mechanicsinitially 8. Although the matrix elements of momentum (p) and position (q) contain the expo-
implied not only that it is meaningless to representelectron motion as nent 27iv,t (where i is the square root of —1, v,is the frequency. associated with the
transition from state i to state k, and t is time), nothing changes .in the calculationsif one
simply writes exp(2zrir;,), without dependence ontime. Actually, Lanczosnotes, in a con-
5. According to Born, one cannotvisualize atomic phenomena ascontinuous processes
sistent discontinuous theory there is no place for such a continuous parameter, and in-
in space (for example, as orbital motion) because spaceitself is not infinitely divisible:
enter eerie eS TIPmR Trreece gr ee

deed the dependence of the matrix elements on timeis never used. Differentiation
there exist ultimate units of matter that cannot be further divided. Born derived this stand with
respect to timeis introducedinto the theory only in orderto preserve the formal analogy
that nature has a scale and that one should not expect to find smaller and smaller ele- with the Hamiltonian equations of classical mechanics: the differentiation of matrix
ments not so much from the presuppositionthat matteris corpuscular as from the special q
with respect to f gives the result dq/dt = (27i /h)(Wq — qW), where Wis the energy
form of quantum laws, in which only whole numbers appear.If physical laws are expres- matrix and h Planck’s constant. This formula indicates a purely algebraic relation between
sionsof relations between whole numbers, more accurate measurement does not add any matrices, rather than an evolutionin time. Pauli presented similar arguments in hisletter
newinformation, and therefore physicists have reached the bottom of nature’s scale (Born to Bohr and speculated that perhaps time could be defined throughthe conceptof energy
1926a, 2). (Pauli to Bohr, 17 November 1925, PC, 260).
alae
22 Chapter Two Matrix Theoryin Flux 23

The meaninglessness of the concept of space-timein the original ver- relied on the “virtual oscillator’ model (Bohr, Kramers, and Slater
=
sion of the matrix theory had momentous consequences. Carried to 1924).1° In order to describe interaction between an atom and radiation
a logical conclusion, it meant eliminating the concept of a particle, or in close analogy with classical theory, the authors replaced the atom
“thinghood,” from the atomic domain. with a set of “virtual” oscillators whose frequencies corresponded to
The original matrix program precluded not only a probabilistic inter- frequencies of transitions between the stationary states of the atom.
pretation (which presupposed that an electron does indeed occupy a They held that in each stationary state the atom continuously radiates
definite position in space and provided formulasfor estimating that po- “virtual” waves whose frequencies correspond to the possible transi-
sition’s probability) but the type of physical interpretation Heisenberg tions from thisstate to all others. The authorsattempted to reinterpret
gavelater in his paper on uncertainty relations (Heisenberg 1927b). To . froma wavetheoretical standpointallthe phenomenasuccessfully ex-
doubtthe existence of the position of the electron in time, as the original plained bylight quanta. Thus Bohr, Kramers, and Slater explained the
theory did, is more radical than to question the existence of simultane- Comptoneffect, in which X-raysscatter and change frequency upon
ous position and velocity. Heisenberg’s uncertainty interpretation does impact with matter, by ascribingthechange in frequency to the Dopp-
notlimit infinite precision of position when momentum remains unde- ler effect. This explanation required them to assumethat the center of
termined: it was a visualizable (anschauliche) interpretation of micro- the waves emitted bythe electron did not coincide with its kinematic
physics, although Heisenberg had previously asserted that such a vis- movement. We have here the first intimation of how incompatible a
ualizable interpretation was impossible. In crder to understand how wave theoretical description is with regular, kinematics.
this change occurred, we need to take into account the overwhelming ~In 1925 Heisenberg joined Kramers in a:paper that dealt with the
success of Schrédinger’s theory and the vehement confrontation that dispersion of radiation by atoms and spelled out, in a rigorous mathe-
decclgped between Schrodinger and Gottingen. matical way, the ideas only roughly outlined in the presentation of
Bohr, Kramers, and Slater: The authors further pursued the program
Wasthe Original Matrix Theory a Theory ofParticles?
of eliminating light quanta from physics. The same kinematic curiosity
The concept of particles was hardly compatible with the original matrix appearedasin the paper by Bohr, Kramers, and Slater, on which it was
approach. The relations between the matrices q, p, and the like were based: the center of the waves emitted by the atom movedrelative to
defined formally so as to retain the Hamiltonian formalist andthe the excited atom (Kramers and Heisenberg 1925).
_ analogy with classical mechanics. The matrix theorists considered the Faced with this incompatibility between kinematics and the wave
positions of electrons themselves to be unobservable: only the fre- theory, the Géttingen-Copenhagen physicists chose the wave theory.
quency, intensity, and polarization of the emitted radiation can be mea- Born, lecturing at MIT in the winter of 1925-26, argued that explana-
sured. As Born announced, “We therefore take from now on the point tions of the Comptoneffect assuming that light quanta and electrons
of view that the elementary waves are the primary data for the descrip- are corpuscular wereless fruitful than explanations relying on the wave
tion of atomic processes; all other quantities are to be derived from approach through the Dopplereffect. Calculations showedthatthe di-
them”(1926a, 70).? rections of motion of the electron and the wave center did not coin-

ProRry
In developing matrix mechanics, the Gertingen physicists tacitly cide: “We therefore stand before a new fact which forces us to decide
_ whetherthe electronic motion or the wave shall be looked uponas the
9. Heisenberg himself pointedly explained the substitution of radiation entities for primary act.” Because “all theories which postulate the motion have
kinematic entities:
proved unsatisfactory,” Born opted for the description based on ele-
In the classical theory the specification of frequency, amplitude and phase
of all the light waves emitted by the atom would be fully equivalentto mentary waves (1926a, 70).
specifying its electron path. Since from the amplitude and phaseof an emit- Heisenberg gradually realized howbasic is the incompatibility be-
ted wavethecoefficients of the appropriate term in the Fourier expansion tween the usual kinematics and the wave theory of light. Even for a
of the electron path can be derived without ambiguity, the complete elec- hydrogen atom (a case of one-particle periodic motion), according to
tron path therefore can be derived from a knowledgeof all amplitudes and
" phases.Similarly, in quantum mechanics too, the whole complex of ampli- 10. Slater’s original conception involved light quanta guided by a “virtual”field; Bohr
tudes and phasesofthe radiation emitted by the atom can be regarded as a’ and Kramersdisposedofthe light quanta. For a discussion of Bohr’s opposition to Ein-
completedescription of the atomic system,althoughits interpretation in the stein’s idea of light quanta andits influence on the Bohr-Kramers-Slatertheory, see Klein
sense of an electron path inducing radiation is impossible. (1933, 292) (1970).
 atreh a nidlanuaat

24 Chapter Two
Matrix Theory in Flux 25

the wave theory one should obtain equidistant spectral lines, and not theoretical (frequency, intensity, and polarizationof radiation), not par-
lines merging into a continuouslimit series (Heisenberg 1926b, 990).!! ticle kinematic (position in space, or evenits probability).
One could adhere either to ordinary kinematics or to the wave theory,
*

but not to both; the recurrent confrontation with kinematic peculiarities Stationary States and Quantum Jumps
made Heisenberg fully aware of the need to choose between the usual THe matrix theorists justified the existence ofstationary states by anal-
kinematics and the wave theory. Heisenberg chose thelatter and re- ogy with waves, very muchin thespirit of Schrédinger’s explanation in
interpreted kinematics in such a waythat it would lead to the correct terms of vibrations, which they later rejected vigorously. One result
spectral, and not the orbital, frequencies of the emitted waves.
the
authors of the Dreimannerarbeit were proud of wasthat in the matrix
This account suggests that the elimination of unobservables wasin- theory the assumption that stationary states with certain discrete en-
voked ex post facto—asjustification and not as a guiding principle—a ergiesexisted was not arbitrary, since their existence followed mathe-
point I will elaborate on in chapter 3. The following statement by Hei- matically from the matrix formalism: “The existence ofdiscrete station-
senberg supports this conclusion: “Onecouldcircumventthis difficulty ary States is just as natural a feature of the new theory as, say, the
only by giving up altogether the assignmentto theelectron or to the existenceofdiscrete vibration frequencies in classical theory” (Born, Heisen-
atom of a definite point in space as a function in time; for justification berg, and Jordan 1926, 322, myitalics). Originally, matrix mechanics
one had to assumethat such a pointalso cannotbe directly observable” and Schrédinger’s theory had more in commonthanis usually appre-
(1926b, 990, myitalics). ciated: both considered somevibration process as primary; but whereas
The Géttingen physicists regarded the virtual oscillator model as Heisenberg reinterpretedclassical space-time throughthese vibrations,
more than a heuristic device. For Born,virtualoscillators were “the real Schrodingerleft the usual concepts of space-time unchallenged. Schré-
primary thing,” and the interaction of electrons in the atom “consists of dinger, indeed, initially regarded his own work as being in the same
a mutual influence (irradiation) exerted by virtual resonators on each vein as the matrix approach. Even though their methods differed, “in
other” (1924, 190). Heisenberg assumed that “something in the atom its tendency, Heisenberg’s attempt stands very near the present one,”
must vibrate with the right frequency” (Heisenberg to van der Waer- he asserted (Schrédinger 1926b, 30).
den, 8 October 1963, van der Waerden 1967, 29). Pauli (1926b) described The authors of the matrix method considered oneresult of the theory
an atom atthe time as a collection of harmonicpartial vibrations, asso- particularly valuable: its ability to obtain discrete solutions without as-
ciated with transitions between different stationary states, and not asa suming a priori that stationary states existed as Bohr had. Heisenberg
constellation of particles tied kinematically to the occupation of certain was especially proud of the independence of the new-quantum me-
stationary states. These vibrations cannot be combined into “orbits.” 2 chanics from this assumption. Thetheoryis instead derived solely from
As Heisenberg put it, “In quantum theory it has not been possible to the commutation relation pq — gp = (h/27i)1 (Heisenberg to Pauli,
associate the electron with a point in space. . .. However, even in quan- 18 September 1925, PC). The stationary states in the matrix approach
tum theory it is possible to ascribe to an electron the emission of radia- werepurely formalartifacts, characterized by the numbers n (or 11,,...,
tion” (1925, 263). ny in the case of f degrees of freedom), which were associated with the

NEE RPS
Thusthe original matrix theory was saturated with wave theoretical definite values of energy and perhaps of angular momentum.Besides
concepts. Originally, matrix mechanics was a symbolic algebraic theory, this, it was not clear whether the states could be given any physical
and the only concepts connected to the physicalsituation were wave interpretation at all, because they wereall “interlocked” by the formal-
eaLasss oases

ism in such a way that one could notassign a varying physical magni-
11. This is what Heisenberg meant by the remarkat the beginning of his reinterpreta- tude to a given stationary state (Dirac 1925).
tion paperthat “the Einstein-Bohr frequencycondition. . . already represents such a com- oo
plete departure from classical mechanics, or rather (using the viewpoint of wave theory) — Stationary states in the new theory were different from the original
from the kinematics underlying this mechanics,that. . . the validity of classical mechanics conception.* In the old theory the sequenceof stationary states wasthat
simply cannot be maintained” (1925, 261-62). ,
12. Moreover, virtual oscillators, as opposed toclassical theory, could not even be 13. The authors of the Dreimannerarbeitstated this clearly when discussing the solu-
looked upon as charged particles. Their “ghostlike” character was revealed by the fact tion of the harmonicoscillator in the new theory. They argued that nothing changes
that for emission oscillators the expression (e*)2/m* (where e* is the chargeof the virtual physically if the sequence of quantum numbers0, 1, 2, 3, .. . is rearranged into a new
oscillators and m™* the mass) is a negative number (Kramers 1924). SEQUENCE No, My, Mz, M3,... such that certain formulasstill hold.
aia ai aaalll

26 Chapter Two Matrix Theory in Flux 27

of increasing values of energy, and this order had a fundamental physi- preted in an indeterministic fashion. Instead, matrix mechanics was
cal significance (of merging continuously into a classical limit). This viewed asa discrete deterministic theory. In this approach the problem
wasnotthe case in the new theory. “The new mechanics presentsitself wasfully determined when the Hamiltonian of the system was known.
as an essentially discontinuous theoryin that herein there is no ques- By,a principal axis transformation that diagonalized the matrix H, in
tion of a sequence of quantum states defined by the physical process, principle one obtained the energy values and the matrices p and q. Ma-
but rather of quantum numbers which are indeed no morethandistin- trix mechanics was constructed by analogy to be “as close to that of
guishing indices which can be ordered and normalized according to any classical theory as could reasonably be hoped” (Born, Heisenberg, Jor-
practical standpoint whatsoever” (Born and Jordan 1925a, 300-301). dan 1926,322). . . am
The matrix theorists realized that the theory’s inability to incorpo- Whataboutthe fact that the matrix elements represented transition
rate the concept of the state of the system was a serious drawback.It probabilities? The Gottingen theorists made no categorical indetermin-
wasclear that an atomic system can exist in certain states for definite istic deductionsfrom it; rather they simply did not exclude the possibil-
amountsof time; therefore, a theoretical description of such states was ity that this was a temporary weakness of the theory, very much the
needed. The ability of Schrédinger’s theory to give a straightforward conclusion Einstein had reached about transition probabilities in his
definition of a stationary state through a wave function was oneof its quantum theoryof radiation in 1917 (Einstein 1917, 76). Pauli, for ex-
advantages, as Hendrik Lorentz pointed out in a discussion of the com- ample, considered that the inability of the matrix theory to determine
parative merits of the two theories (Lorentz to Schrédinger, 27 May exactly when a transition occurs was a weakness: the time of emis-
1926, Przibram 1967, 44), and as Bohr arguedlater in the Comolec- sion of the photoelectron is definitely. observable, and so a satisfac-
ture (see chapter 6). That superiority originally disposed Bornalsoto - tory physical theory should be able to calculateit. A satisfactory physi-
regard wave mechanics as physically more significant than the matrix cal theory should contain no probability concepts in its fundamental
approach. propositions. Pauli was prepared to “pay a high price”to eliminate
In. order to counter Schrédinger, the Gottingen theorists needed a these probabilities (Pauli to Bohr, 17 November 1925, PC, 260). Heisen-
matrix interpretation of the quantum state.-The only way available at berg, under Pauli’s influence, also considered this question in letter to
the time wasto resuscitate Bohr’s original concepts of stationary states Einstein. Should the timesof transition be regarded as observable, and
and jumpingelectrons. Heisenberg collaborated fullywithBohrforthis would matrix mechanics be able to determine them? The matrix theory
purpose. As he admitted years later: “It was extremely important for’ wasstill in such an incomplete (unfertig) stage, complained Heisenberg,
the interpretation to say that the eigenvalues of the Schrédinger equa- that he simply did not know whatstand to take on these questions (Hei-
tion are not only frequencies—they are actually energies. In this way senberg to Einstein, 30 November 1925, Einstein Archive, Hebrew Uni-
of course one cameback to the idea of quantum jumps from.onesta- versity of Jerusalem). ‘
tionary state to the other. . .. But even when weknew this and accepted In accordance with his emphasis on using observable electromagnetic
the quantum jumps, we did not know what the word ‘state’ could variables, Heisenberg (1926b) presented matrix elements as “radiation
mean”(1973, 269). Schrédinger was desperate aboutthe reintroduction . magnitudes”(Strahlungsgréssen), and not as transition probabilities, in

ertt
of “these damned quantum jumps” (quoted in Heisenberg 1967, 103). his Naturwissenschaften paper. He seemedoriginallyto suppressthe sta-
This weird concept seemed alien not only to his theory but to the ma- tistical meaning of matrix elements, expressing his pleasurethat certain
trix approach,originally constructed to dispense with such illustrative - relations formerly deduced from statistical considerations (frequencies
notions. of transitions) could be deduced solely by the mathematical manipula-
tion of “radiation-value tables” (matrices; 1926b, 990). Heisenberg was
Matrix Mechanics and Determinism
nonetheless opento the possibility that for certain questions quantum
In their recollections, the founders of quantum mechanics described theory might not go beyondstatistical answers. Einstein himself was
their efforts to construct the new quantum mechanics as guided by a not sure whether a deterministic account of microscopic phenomena
belief in indeterminism. Historians and philosophers of science often would be possible, or whether there would always remain statistical
follow this lead, seeking the sources of such beliefs in the cultural residue(Einstein to Born, 27 January 1920, Born 1971). The answer was
milieu (Jammer 1966; Forman 1971). Yet we find no strong opinions ex- unclear to the majority of physicists. The use of statistical methods is
pressed on theissue of causality during the creative stages of the erec- ‘not, of course, in itself evidence of a desire to disposeof causality;it
is
tion of the new theory. Nor was the new matrix theory initially inter- rather a sign that, at least for the time being, statistical methods yield
28 Chapter Two
Matrix Theory in Flux 29

the best scientific results. Heisenberg’s preference for indeterminism impasse: the solution of the hydrogen atom could not lead to any fur-
evolved gradually in complex theoretical and human dialogical con- ther advance ofthe theory, and consequently the perturbation methods
texts (see chapters 4, 5, 9, and 10). discussed in the Dreimannerarbeit were useless for new physical appli-
Mathematical Difficulties of the Matrix Approach cations. This serious drawback was duly noted, by Fritz London
, for
example (London 1926, 921).
The matrix method of solving quantum mechanical problems wasfor- The authors of matrix mechanics found themselves in an ironic
mulated in an astonishingly simple and general way. For anypair of posi-
tion. They had erected the new quantumtheory primarily in order
values po, qo that satisfies the basic commutation relation P9I- 9p = to
approach multiple-electron systems, where the old methods had al-
(h/277)1 (as, for example, in the simple case of the harmonic oscillator), ready failed: Yet the mathematical difficulties of matrix mechanics vir-
solving the quantum problem (the problem of integrating the canonical tually prohibitedits extension beyond simple systems with one moving
equations for the given H(pq)) reducesto determining a matrix S that particle. In fact,it was only after Schrédinger’s wave mechanics had
diagonalizes the Hamiltonian: (pq) = SH(poqo)S~! = W, where W is appearedthat Pauli succeeded in calculatingthe intensities of the Bal-
a diagonal matrix. $ then givesthe solution of the canonical equations mer terms, and onlybyusing Schrédinger’s eigenfunctions of the hy-
p = SpoS "and gq = Sq,S~. drogen atomforhis solution (Pauli to Landé, 2June 1926, PC, 327). The
The simplicity of this economical and beautiful formulation was, superiority of Schrédinger’s solution of the hydrogen atom over those
however, quite misleading: it was practically useless becauseof the dif- © _ of Pauli, Dirac, and Wentzel showedthe clear advantage of his theory
ficulty of calculating the reciprocal matrix S-!. Matrix mechanics could for solving quantum systems in centralfields. Dirac’s and Wentzel’s cal-
not solve any general case, only, by perturbation methods, those physi- - culations did not determine whether the quantum numbers of Balmer
cal problemsthat could be considered approximationsto solved cases. energy terms wereintegers or half-integers, while Schrédinger’s calcu-
The application of perturbation methods presupposed “that several lations resulted unambiguouslyin half-integers. Both Dirac and Went-
specially simple systems which are used as starting points in calculus zel, and later Heisenberg andJordan, used a two-dimensional treatment
are completely known’(Born 1926a, 99; see also London 1926). The key instead of the correct, three-dimensional one (Schrédinger to Lorentz,
wordhere is “completely”: the “interconnectedness” of the ‘quantum... 6 June 1926, Przibram 1967, 64-65; Heisenberg and Jordan 1926).'* They
system in the matrix approach was such that onehad to know boththe reached agreementwith the experimental results bytailoring question-
frequencies and the intensities of the unperturbed system in order to able methodsto results knownin advance:if they had conducted the
calculate any of them for the perturbed system. calculations rigorously, they would have obtained the wrong answer,
The crucial test of the theory was whether it could solve the central as Schrodinger suspected and John van Vleck (1973) has since demon-
problem of the atomic domain—the hydrogen atom. No wonder Hei- strated. Dirac himself used Schrédinger’s solution in The Principles of
senberg rejoiced when Pauli succeeded in working out the Balmerfor- Quantum Mechanics (Dirac 1930), and Pauli scolded Born and Jordan for
mula for the hydrogen spectrum, though Pauli did so with difficulty including his own complicated solution, inferior to Schrédinger’s, in
andat the price of additional assumptions (Pauli 1926a; Heisenberg to their book Elementare Quantenmechanik (Born and Jordan 1930). There
Pauli, 3 November 1925, PC). P. A. M. Dirac and Gregor Wentzelalso were other substantial mathematical difficulties as well, including the
obtained the solution of the hydrogen atom, a success for the matrix problem of coordinates (for a full discussion, see Beller 1983).
approach, but not the grand triumph generally assumed (Dirac 1925,. The matrix approach nevertheless showed promise. It seemed espe-
1926; Wentzel 1926a). In order to proceed from the simple hydrogen cially well suited to calculations involvingangular niomentum matrices.
atom to more complicated atomic systems by perturbation methods, Consequently, it allowed straightforward quantization of the angular
one needed to know both the frequencies and the intensities of radia- momentum components(thatis, the calculation of magnetic quantum
tion of the hydrogen atom; but Pauli, Dirac, and Wentzelall failed numbers) and the deduction ofselection rules without extraneous as-
to calculate the intensities. Moreover, it seemed hopeless to obtain sumptions (unlike the old quantum theory, where certain orbits, lead-
this solution by matrix methods.'4 Matrix mechanics was thus at an ing to collision between electrons and the nucleus, were excluded). It

14. This opinion is expressed by Jordan:-“It appeared hopeless to calculate the matrix 15. ‘Schrodinger initially obtained hailf-integers, not integers, because spin was
not yet
elements” (1927a, 641). taken into account.
ent t|

30 Chapter Two Matrix Theory in Flux 31

also enabled the intensities in the normal Zeeman effect and the Stark changing moodsin the fall of 1925 reflected his doubts about the new
effect to be calculated (Born 1926a; Born, Heisenberg, andJordan 1926). theory. Thus, in a letter to Pauli, Heisenberg wrote that he considered
Moreover, matrix mechanics successfully treated the harmonic and an- the principal axis transformation the most importantpart of the whole
harmonicoscillators, dispersion, and, though incompletely, the hydro- theory (Heisenberg to Pauli, 23 October 1925, PC), only to call it “for-
gen atom. mal rubbish” three weeks later (Heisenberg to Pauli, 16 November 1925,
PC). Heisenberg’s letters to Dirac and Einstein at the time do notreflect
The Emotional Confrontation between the a belief in the validity of the matrix approach. “I read yourbeautiful
Matrix Physicists and Schrodinger work with great interest,” he wrote to Dirac. “There can be no doubt
that all yourresults are correct, insofar as one believes in the new theory’
The successes of the matrix mechanical method were modest compared (quoted in Dirac 1977, 124, myitalics).1°
with the difficulties encountered. Most of the problemstreated using Nordothe original published versions of matrix mechanics convey
the matrix approach had been solved previously. Their solutionwas anyfeeling of an extraordinary scientific breakthrough.A cautious tone
therefore more an encouraging sign of the newtheory's possible va- is adopted in the Dreimannerarbeit: even though the authors would
lidity than a guarantee of future success. The most outstanding atomic like to concludethat the theory might be the correct one becauseofits
problems—hydrogen intensities and helium energy terms—awaited mathematical simplicity and unity, they realize nonethelessthatit is yet
solution. They proved to be as resistant to the matrix approach as to unable to solve the crucial problemsof the quantum domain (Born, Hei-
the old Bohr-Sommerfeld theory. Even those who favored the matrix senberg, and Jordan 1926). Even less optimistic are Born’s concluding
approach were reluctant to use it. Thus Arnold Sommerfeld declared: remarks in his MIT lectures: “Only a further extension of the theory,
“There wasclearly an element of truth in it [matrix mechanics], butits which inall likelihood will be very laborious, will show whetherthe
handlingis frighteningly abstract” (1927,.231). principles given aboveare really sufficient to explain atomic structure.
There were also grave problemsof physical interpretation. “Heisen- Evenif weare inclined to putfaith in this possibility, it must be remem-
berg’s theory in its present form is not capable of any physical interpre- bered that this is only the first step towardsthesolution ofthe riddles
tation at all” was the harsh verdict (Campbell 1926, 1115). There was of the quantum theory” (Born 1926a, 128). go
‘'sinaply no’space-time “gravy,” to use Hermann Weyl's characterization, It is not surprising, therefore, that except for the inventors of the
in the matrix representation (Sommerfeld 1927, 231). The elimination of new theory and those under their direct influence, only a very few
unobservables led to the elimination of space-time and of physical re- physicists attempted to employ the matrix method. Most physicists, it
ality itself, including the by then familiar electrons. The extreme Mach- seems, decided to “wait and see’”— they would not submitto studying
ean approach, which eliminated everything but immediate sense per- complicated mathematical techniques until the new theory proved its
ception (intensities, frequencies, and polarization of spectral lines) and worth. Nor werethey inclined to direct their students to study matrix
a highly abstract uninterpreted formalism, did not seem tenable. Ein- mechanics. Felix Bloch, Peter Debye’s student, was not even aware of
stein questioned the soundnessofit, and London (1926) and Nicholas the matrix theory before he learned aboutit from Schrédinger’s publi-

EMT
Rashevsky (1926) argued that it was inconsistent. The ability to under- cation.”” Eventhe biggest success of the matrix approach—thesolution
stand the transition from the micro- to the macrodomain was ques- of the Balmer termsfor the hydrogen atom—lackedboth the complete-
at

tioned. Nor could the matrix approachinitially be reconciled.with the - ness and the elegance one would expect from a full-fledged theory.
3 definitionof a quantum state. Thelack of visualizability was a heuristic, Pauli himself was pointing to the limitations of the matrix approach
if not a conceptual, hindrance. To build a truly discontinuous theory, whenlater referring to his own result as having been derived by an
| one had to proceed with no suitably developed mathematics and in a “inconvenient and indirect method”(1932, 602).
a virtually complete conceptual vacuum. After the emergence of Schré- In contrast, Schrédinger’s theory, mathematically powerful and
&
dinger’s theory, it is mo wonder the approach was abandoned, first
by Born (who proposed the probabilistic interpretation), then by Pauli 16. Heisenberg’s insecure and self-critical mood obviously persisted after Pauli’s solu-
(who extended Born’s approach), and later by Heisenberg (who ad- tion of the hydrogen spectrum, which, according to the usual historical accounts, dis-
pelled any doubts about the correctness of the newtheory.
vanced an interpretation of the indeterminacy relations).
17. As he wrote yearslater: “I did not learn about the matrix formulation of quantum
No wonder, too, that the creafors of the new mechanics did not seem mechanics by Heisenberg, Born and Jordan until I read that paper of Schrédinger, in
to have much confidence in their theory in the beginning. Heisenberg’s which he showed the two formulations to lead to the sameresults” (Bloch 1976, 24).
 Matrix Theory in Flux 33
sc povskneaPRRET

32 Chapter Two
mechanics, so
com- in the beginning because there already was one quantum
familiar (as well as physically more accessible), was hailed by the no other was needed. Dirac admitted that “he defini tely had a hostility
amongall
munity of theoretical physicists. “It was the most astonishing to Schrédinger’s ideas to begin with, which persis
ted for quite a while”
feld
the astonishing discoveries of the 20th century,” declared Sommer (1977, 131).
It is a historic al myth that the enthus iasm —he éonsid-
(quoted in Moore 1989,2). Pauli also reacted to Schrédinger’s ideas with suspicion

ue
the conserv ative part of the scientif ic commun ity. It was Pauli to Sommer-
waslimited to
Dar- ered Schrédinger’s approach verriict (crazy, foolish;
welcomed,for example, by such unconventional minds as Charles feld, 9 February 1926, PC). Learningof Sommerfeld’s
regard for Schr6-
: oo at it, decided thatit
win, Fritz London, and Enrico Fermi. dinger’s work, however, Pauli took a closer look
way,
The success of Schrédinger’s theory contributed, in an indirect belonged amongthe “most meaningful” recent publications
, proved the
learned
to the dissemination of the matrix approach. Many physicists equivalence between matrix mechanics and wave
mechanics, and sub-
-
about the existence of matrix mechanics from Schrédinger’s publica sequently stroveto elucidate its physical meaning bytry
ing to under-
tions. The proofof the equivalence of matrix and wave mechani cs. en- inger’ s forma lism and Einstein-
ity. stand the connection between Schréd
dowed the unfamiliar and abstract matrix approach with credibil de Broglie waves (Pauli to Jordan , 12 April 1926, PC).
d to Schréd inger’ s theory with — proof
the equivalence
Yet the young matrix theorists reacte Heisenberg, who could no longer contend after
t for Born; see the next sectio n of this ed that its only value was
disbelief and hostility (excep
s theory was that Schrodinger’s theory was wrong, declar
chapter). Heisenberg hoped initially that Schrédinger’ its ability to calculate the matrix elemen ts (Heise nberg to Pauli, 8 June
rately ignored
wrong. Years later Heisenberg explained that he delibe 1926, PC). Jordan (1927a) arrive d at a similar conclu sion: the ‘“meaning
alenc e paper becaus e of y provided a new mathe-
wave mechanics before Schrédinger’s equiv
5 AFLGators =

with Heisen - of the thing” was clear—eigenfunctions merel

the interpretation Schrédinger attach ed to it (inter view matrix mechanics. Jordan
s initial © matical method for solving the equations of
berg, AHQP). This explana tionis not convin cing. Schréd inger’ nts of Schrédinger’s inter-
atical became one of the most militant oppone
presentation of his theory was undertaken. in “neutral mathem pretive efforts. Schrodinger should have beensa tisfi ed with the mathe-
form” (Schrédinger 1926a, 9). Schrédinger’s opinion aboutt
he need “to not have even attem pted anything more
atom, which matical advance and should
connect the function y with somevibration process in the than a mathematicalelaboration, declared
Jordan (1927a, 615). The new
ity than the electr onic orbits, the real analogy with the older
would morenearly approachreal theory must be “interpreted physically in close
is being very much questi oned to-day ” (1926a, 9), and with Heisenberg’s
existence of which notions of stationary states and quant um jumps,
berg’sr einter - s collected papers on
did not contradict the original inspiration behind Heisen theory” (1927b, 567). In a review of Schréd inger’
reality of electr on or- theoryto be devoid of
pretation paper. Heisenberg also doubted the wave mechanics, Jordan declared Schrédinger’s
ce radiation that such was the opin-
bits and tried to detect those internal vibrations that produ any physical meaning and stated (incorrectly)
in agreement with experiments. There was no substa
ntial reason for physic ists (Jordan 1927d). A
inger be- ion prevailing among the majority of s work with-
Heisenberg’s aversion to Schrédinger’s theory; indeed, Schréd novice in the field should not be expose d to Schréd inger’
“tende ncy” were
lieved initially that his and Heisenberg’s work in their out prior appropriate instruction in physic
al matters in the Gottingen-

OU SET
inger’ s theory seems more inger, understandably, was
very close.’ Heisenberg’s hostility to Schréd Copenhagen spirit, Jordan continued. Schréd
anybody to Born (Schro dinger
likely to be connected with his instinctive reluctance to admit outraged, and he complained about this review
his own.
else into territory that the ambitious Heisenberg considered to Born, 6 May 1927, AHOP). Born
admitted that Jorda n somewhat
This probably wasalso the reason for Dirac’s initial opposition
to Schro-- youthful temperament
exceeded his limits and blamed it on Jordan’s
dinger’s theory. As Dirac reportedlater, he ignored Schrédinger’s theory (Born to Schrédinger, 16 May 1927, AHQP).
self-sufficient physi-
the way in which Schrédinger’s hopethathis theory would havea
18. Lorentz wrote to Schrédinger: “I was particularly pleased with expectation wasconsistent
and show that these satisfy the equations of cal interpretation was not unreasonable. This
you ... construct the appropri ate matrices
Einstein, that mathema-
motion. This dispels a misgiving that works of Heisenberg,
Born andJordan. .. inspired with the belief of physicists from Galileo to
akable signs of a theory’s physical
in me” (Przibram 1967, 43). tical simplicity and power are unmist
19. The hint abouta possible “beat” wave model could have aroused
Heisenberg’s ob- attemp ted initial ly to elucidate some
able realistic models. Yet
significance.*° Schrodinger
jection because he no longerinclined toward building visualiz
al and understated that feelings of many physicists when he
the interpretive issues in Schrédinger’sfirst paper were so peripher 20. Schrédinger undoubtedly expressed the
overlook them instead of the advance Schrédinger nce implie d by the elimination of intuitive space-time
it would have been more natural to pointed to the conceptual hindra
had made.
 34 Chapter Two

Matrix Theory in Flux 35

“conceivable mechanism” by which micr
ophenomena take place in
regular space-time. Physical reality, accordin denounced, as a rich source of suggestions for const
g to Schrédinger’s early at- ructing a rival in-
terpretation. Heisenberg’s elaboration of the
uncertainty relations
hinged on a translation of a “wave packet a la Schré
dinger” (Heisen-
mechanical equations. This ontology soon prov berg to Pauli, 23 February 1927, PC) into the langu
ed to be abortive ; yet the ageof a particle on-
“intuitiveness” of Schrédinger’s approach tology (see the detailed discussion in chapter
ability to provide (ornot) classical was not confined to its 4). Bohr understood
idealized space-time models. clearly that the wave theory played an essent
the physical significance of Schrédin Rather, ial role in Heisenbere’s
ger’s theorylay in its ability considerations, both historically and philosophi
to pro- cally.“ Bohr himself
vide some qualitative handle on the essential aspec strove to elucidate the physical significance
Schrodinger’s theory deciphered ts of the microworld. of Schrédinger’s wave
the mystery of quantization; it theory in his Como lecture(see chapter 6). Like
ex- the majority of the sci-
plained why atoms in stationary states do entific community(the alleged “conservatives”
cial line that denied physical significance notra diate. Despite the offi- ) he preferred the intui-
tiveness of Schrédinger’s version to the formal abstra
matrix physicists, even Jordan, could to Schré dinger’s theory, the ction of the matri
x
theory: “This ingenious attack [the matrix appro
Schr
not help but sometimes praise ach] upon the problem
édingerfor just these assets
21 of the quantum theory makes, however, great demands
on our powe r
Heisenberg understood early that of abstraction, and the discovery of newartific
there must be some very close con es [wave mechanics]
nection between Schrédinger’s app - which, in spite of their formal character, more closel
roach andhis own. Herealized tha y meet our de-
it was Schrédinger’s theory that might t mandsfor visualization, has therefore been of profo
provehelpfulin elucidati
ng the undsignificance in
physical meaning of his own abstra the developmentandclarification of quantum mechanics”
ct version of quantum mechanics
. (Bohr 1929a,
110-11).
“A few weeks ago anarticle by Sch Noris the emergence of the G6ttingen-Copenhagenstatist
rédinger appeared .. . whose con- ical ontol-
tents to my mind should be closely con ogy in particle terms comprehensible without taking
nected with quantum mechan- into consideration
ics. Haveyou considered how far Schrédinger’ the selective borrowing of Schrédinger’s interpretive
gen atom is connected with the qua s treatmentof the hydro- ideas and_ their
ntum mechanical one? This mathe- translation into particle language. Born’s statistical
interpretati
on was,
matical problem interests me especially because . .. one to a large degree, a translation into particle language
can win from ita of Schrédinger’s
eat deal for the physical significance idea that the y-function determines the electron charg
of the theory (Heisenberg to Dira e density; Pauli’ -
9 April 1926, AHQP, myitalics). Pauli agreed,ci c, fused Schrédinger’s idea of a “weight function” with
Born’s interpreta-
“jook[ing] at the problem from two diff ting the possibility of tion, arriving at the interpretation of the wave
erent sides” (Pauli to Heisen- function as givin g the
berg, 12 April 1926, PC). probability that the system is ina specific configuratio
n. Dirac similarly
Heisenberg fully exploited the inte used the suggestiveness of Schrédinger’s conception
rpretive possibilities opened s in forming his
bythe wave theory, using Schrédinger’s imagery, probabilistic ideas.*4
which he publicly
pictures in matrix mechanics. The prob 22. Heisenberg wrote to Dirac about his confrontatio
lems that
atomic physics had to treat theoretica n with Bohroverthe uncertainty
were presented, Schrodinger Poin lly
ted out, in “an eminently paper: “Prof. Bohr says, that onein all those exampl
ample, how twocolliding atoms or mole intuitive form; as, for ex- essees the very important role,which
cules reboundfrom one another, the wave-theory plays in my theory and, of course,
or how an elec- he is quite right” (Heisenberg to Dirac,
tron or a-patticle is diverted whenit js shot 27 April 1927, AHQP).
through an atom with a given velocity
with the initialpath at a given Perpendiculardistancefr and 23. Bohr expressed similar reservations about
om the nucleus.” How is one even the mathematical complexity of matrix
to begin to treat such problems if one Operates mechanics and maintained his preference for
tion probabilities, energy levels, and the only with such abstract ideas as transi- the wave mechanical version throughout
like? Darwin's wordsthat “the ultimate his life. In 1938 he wrote: “On accountof the intrica
will be one of space and time again” theory te mathematical operations involved,
echoed Schrédinger’s hopes (Dar it was, however, of utmost importance, not
AHQP). > win to Bohr, 1928, only for the practical use of the formalism,
but evenfor the elucidation of essential aspect
s ofits consequences,that the treatmentof
any quantum-mechanical problem could be shownt
o be essentially reducible to the so-
lution of a linear
differential equation” (Bohr 1939, 387).
24. “Schrédinger’s wave representation of
quantum mechanicsha s provided new ways
of obtaining physical results from the theory
based onthe assumption that the square of
the amplitude of the wave function can
in certain cases be interpreted as probabi
(Dirac 1927, 621). lity”
 36 Chapter Two
Matrix Theoryin Flux 37
From a purely conceptual point of view, the intens
e hostility to far-reaching consequences. It implied no less than the powerto “in-
Schrédinger’s ideas (rather than, say, polite disagreement)is
puzzling.*> fluence . . . the research of the following century” (1952, 60). The “vic-
Partly, as I have noted, the matrix theorists resented
Schrédinger’s in- tory” in Bohr’s and Schrédinger’s characterization similarly meant “to
trusion into their “territory.” It did not seem feasible that
an outsider realize one’s wishesfor the future of physics.” ”” The weightof the threat
could arrive at the long-sought solution of the quantu
mriddle, and that

tae
the solution could be that simple. Schrédinger’s success to the Géttingen-Copenhagen version of physics was revealed at the
aroused both Munich conference held at the end of summer 1926. Most participants
disbelief and envy. The assessment by Heisenberg thatit was “too good |
to be true” reveals this attitude (1971, 72). So does Born’s dismissal: “It
there aligned themselvesnot only with Schrodinger’s methodsbut also
_ with his interpretive aspirations. Even Sommerfeld, or so it seemed to
would have been beautifulif you [Schrédinger] were right.
that beautiful happens, unfortunately, seldom in this Somet hing Heisenberg, succumbedto the persuasive force of Schrédinger’s mathe-
Schrédinger, 6 November 1926, AHOP).
world” (Born to matics.** Heisenberg’scritique of Schrédinger’s intentions “failed to im-
. Se
Yet disbelief and envy, strong as these feelings are, Ry Be press anyone,” and Wilhelm Wien crowedto the despairing Heisenberg
were only a part that his version of quantum mechanics with the nonsensical quantum
ofthe psychosocial setting in which the new quantum
mechanics was jumping was “finished” (Heisenberg 1971). Heisenberg immediatelyre-
erected. The Géttingen-Copenhagen physicists saw in Schrédinger’s
program something capableof extinguishing the matrix appro ported this alarming situationin a letter to Bohr, and Schrédinger was
ach(in- subsequently invited to Copenhagen for what would become heated
terview -with Jordan, AHOQP). Had Schrédinger succeeded in giving
a satisfactory interpretation without the Géttingen-Copenhagen con- discussions.”” After Schrodinger left Copenhagen, a feverish huntfor an
cepts, he might haveeliminated the need for the matrix approachalto- adequate interpretation occupied both Heisenberg and Bohr. .
gether, in view of the mathematical equivalence of the matrix and wave The emerging Géttingen-Copenhageninterpretation did not weaken
the “victory” of Schrédinger’s methods. Even physicists who had made
theories and the greater manageability and familiarity of the latter.
Many physicists found Schrédinger’s theory much moreattractive and someinitial contributions to the matrix approach turned to the wave
wereinclined to regard Schrédinger’s physical approach as the theory with relief, Wentzel and London among them. Results obtained
more in the matrix framework were often translated into wave language,
correctof the two candidates (interview with Jordan, AHPQ).
as, for example, William Gordon's wave treatment of the Compton
A flood of papers, most following Schrédinger and ignoring the ma-
effect, which came after Dirac’s work in the matrix framework. Such _
trix approach,highlighted the reality of the threat. Schrédinger’s theory
was successfully applied to a great variety of problems unamen duplication of scientific results was justified by the argument that
able to “Schrédinger’s methods possess the advantage of using the familiar
matrix treatment. Born recalled the successes of Schrédinger’s mechan-
mathematical forms only” (Gordon 1927, 117). When Born tried to
ics: “In the meantime, Schrédinger’s wave mechanics appeare
d, and show that someresults achieved in the wave framework could also be
won the approbation oftheoretical physicists to such an extent that
our.own matrix method was completely pushedinto the backgro easily obtained by matrix methods, Paul Ehrenfest called the latter “a
und, bad habit” (interview with Oskar Klein, AHQP). Matrix tools could not
particularly after Schrédinger himself had shown the mathematical
equivalence of wave and matrix mechanics” (1971, 104).
Expressing
himself more strongly, Born said during an interview with Thomas 27. Schrédinger to Bohr, 23 October 1926, AHQP;reprinted in BCW, 6:459~61, trans-
Kuhn: “Wave mechanics was considered the real quantum mechanics. lation on 12.
by everybody, while matrix theory was completely neglected”(in 28. Sommerfeld did have somereservations but apparently abstained from expressing
ter- them. Sommerfeld reported to Pauli about the Munich conference in a letter: “ ‘Wave
viewwith Born, AHOQP).* Born was indeed unhappyat the time about mechanics’ is an admirable micromechanics,yetit is still far off from solving the funda-
the “world-wide victory” of wave mechanics (Schrédinger to Born, mental quantum riddle” (Sommerfeld to Pauli, 26 July 1926, PC).
17 May 1926, AHOP), - 29. Heisenberg’s description of these discussions is among the most quoted passages
This “victory,” as Heisenberg clearly understood, was pregnant with in the literature of the history of quantum theory. Heisenberg’s tale is highly dramatic:
“Though Bohr was an unusually considerate and obliging person, he was able in such a
_ ®.Heisenberg used such emotionally charged wordsas abscheulich discussion, which concerned epistemological problems which he considered to beofvital
ing) in characterizing Schrédinger’s approach. The connectio (repelling, disgust- importance, to insist fanatically and with almostterrifying relentlessness on complete
n between anecdotal and
conceptual history, and the formative fole of emotionsin cognitive endeavors clarity in all arguments. . . . He would not give up, even after hoursof struggling. . . . It
,is analyzed was perhaps from over-exertion that after a few days Schrodinger becameill and had to
in Beller (1996b).
26. Citation analysis confirms Born’s judgment(Kojevnikov and Novik 1989). lie abed as a guest in Bohr’s home. Evenhereit was hard to get Bohr away from Schré-
dinger’s bed” (1967, 103).
38 Chapter Two Matrix Theoryin Flux 39

compete with wave mechanical ones. By 1929 “the research was domi- occupy importantchairs of their own (Cassidy 1992). Young physicists,
nated by wave mechanics, and matrix mechanicsso to speak only came who streamed into these centers from all over the world, were exposed
back through group theoretical arguments” (interview with Hendrik automatically to the new philosophy. Because they were more inter-
Casimir, AHQP). Born and Jordan madea last, desperate attempt to ested in calculating and obtaining definite scientific results than in phi-
opposethis trend: they wrote a book— Elementary Quantum Mechanics losophizing, most of them simply adopted the official interpretation
(1930)—relying solely on matrix methods, in which Schrédinger’s wave without deep deliberation. Heisenberg was awareof this. In his 1930
functiondid not appear even once. This book “in view of the general book, which he dedicated to the “diffusion of the Copenhagenspirit,”
predisposition in favor of SchrGdinger. . . was not favorably received” Heisenberg conceded that “a physicist more often has a kindoffaith
(interview with Born, AHQP). Pauli (1932) himself wrote a devastating in the correctness of the new principles than a clear understanding of
review: them” (1930, preface).
The aging Schrédinger witnessed a remarkable state ofaffairs: the
“What is the “elementary” quantum mechanics of the present vol- universaluseof his theory coupled with an almosttotal rejection of his
ume? ... Elementary is that quantum mechanics which makes use of
interpretation. Schrédinger’s methods proved indispensable. His phi-
elementarytools, and elementary tools are purely algebraic ones; the
use of differential equations is... avoided as muchas possible. . .
losophy did not.
Manyresults of quantum theory can indeed notbe derived at all with
the elernentary’ methods defined above,-while the others can be de- Born’s Probabilistic Interpretation:
rived only by inconvenient and indirect methods. Amongthe latter _A CaseStudy of “Concepts in Flux”
results belong, for instance, the derivation of the Balmer terms, which
is carried out-in matrix theory accordingto anearlier paperof Pauli’s Born’s probabilistic interpretation of the wave function occupiesa cen-
dealing with it. In this regard, he will not be able to accusethe re- tral place in the philosophy of quantum mechanics. In the orthodox in-
viewerthat he finds the grapes to be sour because they hang too high terpretationit signifies the abandonmentof determinism.and the intro-
for him.Therestriction to algebraic methodsalso often inhibits insight ductionto a new kind of reality, abstract and “ghostlike.” Accordingto
into the range and the innerlogic ofthe theory. . . . The setup of the Born’s reminiscences, his belief in particles rather than waves,together
bookas far as printing and paperare concernedis splendid.
with Einstein’s idea of a connection betweentheintensity ofthe electro-
Schrédinger’s methods did indeed win an overwhelmingvictory. But magnetic field and the density of light quanta, made it “almostself-
theinterpretation that most physicists seemed to accept wasthe one given understood”to interpret ||?dx as the probability density of particles
by Schrédinger’s opponents. The Géttingen-Copenhagen line was, of (Born 1961). However, scholars have expressed some doubts aboutthe
course, not unassailable—thatcritical debates on the philosophy of authenticity of Born’s recollections.*°
quantum physics continue to the present day attests most eloquently to [have suggested a revision of the history of Borr’s probabilistic inter-
this fact. Yet quite apart from the philosophical question of the validity pretation along the following lines (Beller 1990):
of the opposing approaches, wecan ask a sociological question con- 1. Born’s probabilistic interpretation was a conceptual contribution
cerningtheinitial distribution of forces. We will not be surprisedto find that crystallized over a considerable period of time—time during
out whoprevailed. At first Schrodinger had the emotional support of which Born'sideas, as a result of his dialogues with Schrodinger, Hei-
manyphysicists, some distinguished and somenot. Yet he struggled senberg, Pauli, and other physicists, underwent significant changes.
with the problems of interpretation virtually alone. Those who iden-
tified with his aspirations hardly did anything to advancehis efforts. 30. Pais (1982) hasnoted thatif Born hadreally been stimulated by the analogy of light
They hoped,as Jordan recalled, that Schrédinger would accomplish his intensity as a quadratic function and Schrédinger’s wave function, he would not have
task unaided (interview with Jordan, AHQP). been able, howeverbriefly, to supposeinitially that w rather than |u|? was a measure of
The Géttingen-Copenhagenphysicists, in contrast, presented a united probability, as he didin his first collision paper. Stachel (1986) has argued thatthe source
of Born’s inspiration was not Einstein’s unpublished speculations but rather his work on
front. They cooperated intimately, and each contributed extensively to monatomic gases, in which he suggested a connection between material particles and de
the emergence of the new philosophy. Bohr and Born headed the era’s Broglie fields. Wessels (1981) has pointed out that in none of Born’s original papers on
most prestigious schools of theoretical physics, and promising physi- this subject did heinterpret |y|?dx as a probability of position. These valuable insights
cists were eager to work with them. Soon Pauli and Heisenberg would found additional reinforcement and explanation in my account(Beller 1990).
 38 Chapter Two
Matrix Theory in Flux 39
compete with wave mechanical ones. By 1929 “the
research was domi- occupy important chairs of their own (Cassidy 1992). Young physicists,
nated by wave mechanics, and matrix mechanics so to speak
only came whostreamedinto these centers from all over the world, were
back through group theoretical arguments” (inte exposed
rview with Hendrik automatically to the new philosophy. Because they were more inter-
Casimir, AHQP). Born and Jordan made
last, desperate attempt to ested in calculating and obtaining definite scientific results than
oppose this trend: they wrote a book— Eleme in phi-
n tary Quantum Mechanics losophizing, most of them simply adopted the official interpretation
( 1930) —relying solely on matrix methods, in
which Schrédinge
r’s wave without deep deliberation. Heisenberg was aware ofthis. In his 1930
function did not appear even once. This book
“in view of the general book, which he dedicated to the “diffusion of the Copenhagenspirit,”
predi sposition in favor of Schrédinger .. . was not
favorably
received” Heisenberg conceded that “a physicist more often has a kind of faith
(interview with Born, AHQP).Pauli (1932) himse
lf wrote a devastating new
-in the correctness of the new principles than a clear understanding of
review:
them” (1930, preface). ae
Whatis the “elementary” quantum mechanics
of the present vol- The aging Schrédinger witnessed a remarkablestate of affairs: the
ume? .. . Elementary is that quantum mechanics universal use of his theory coupled with an almost total rejection of his
which makes useof
elementary tools, and elementary tools are interpretation. Schrédinger’s methods proved indispensable. His phi-
purely algebraic ones; the
use of differential equations is... avoided as much losophydid not.
as possible... .
Many results of quantum theory can indeed notbederiv
ed atall with
the elementary methods defined above, while the
others can be de- . Born’s Probabilistic Interpretation:
rived only by inconvenient and indirect methods.
Among thelatter A Case Study of “Concepts in Flux”
results belong,for instance, the derivation of the Balme
rterms, which °
is carried out in matrix theory according to an earlie
r paperof Pauli’s Born’s probabilistic interpretation of the wave function occupies a cen-
dealin g with it. In this regard, he will not be able
to accuse the re- tral placein the philosophy of quantum mechanics. In the orthodox in-
viewerthathefinds the grapes to be sour because they
hangtoo high: terpretationit signifies the abandonmentof determinism andthe intro-
for him. Therestriction to algebraic methodsalsoofte
ninhibits insight
into the range and the inner logic of the theory.
... The setup of the
duction to a new kindofreality, abstract and “ghostlike.” According to
book asfar as printing and paper are concerned is splendid.° Born’s reminiscences,his belief in particles rather than waves, together
.
with Einstein’s idea of a connection between the intensity of the electro-
Schrédinger’s methods did indeed win an overw
helming victory. But magnetic field and the density of light quanta, made it “almost self-
the interpretation that most physicists seemed to accep
t wastheone given understood”to interpret |¢|?dx as the probability density of particles
by Schrédinger’s opponents. The G6ttingen-Copen
hagen line was, of (Born 1961). However, scholars have expressed some doubts about the
course, not unassailable—that critical debates
on the philosophy of authenticity of Born’s recollections.°
quantum physic s continue to the present day attests most eloqu
this fact. Yet quite apart from the philosophical questi ently to I have suggesteda revisionof the history of Born’s probabilistic inter-
onof the validity pretation along the followinglines (Beller 1990):
of the opposing approaches, we can ask a sociological
question con-
cerningthein itial distribution of forces. We will not be 1. Born’s probabilistic interpretation was a conceptual contribution
surpr
ised to find
out whoprevailed. Atfirst Schrédinger had the that crystallized over a considerable period of time—time during
emotional support of
many physicists, some distinguished and which Born’sideas, as a result of his dialogues with Schrédinger, Hei-
somenot. Yet he struggled:
with the problemsof interpretation virtually senberg, Pauli, and other physicists, underwent significant changes.
alone. Those who iden-
tified with his aspirations hardly did anything to
advance his efforts.
They hoped, as Jordan recalled, that Schrédinger 30. Pais (1982) has noted thatif Born hadreally been stimulated by the analogy oflight
would accomplishhis intensity as a quadratic function and Schrédinger’s wave function, he would not have
task unaided (interview with Jordan, AHQP
). been able, howeverbriefly, to suppose initially that y rather than |y|? was a measureof
The Gottingen-Copenhagen physicists, in contr
ast, presented
a united probability, as he didin hisfirst collision paper. Stachel (1986) has arguedthat the source
front. They cooperated intimately, and each
contributed extensively to of Born’s inspiration wasnot Einstein's unpublished speculations but rather his
work on
the emergence of the new philosophy. Bohr and monatomic gases, in which he suggested a connection between material particles and de
Born headedthe era’s
mostprestigious schools of theoretical physi Broglie fields. Wessels (1981) has pointed out that in none of Born’s original papers on
cs, and promising physi- “this subject did he interpret |w|?dx asa probability of position. These valuable insights
cists were eager to work with them. Soon
Pauli and Heisenberg would found additional reinforcement and explanation in mvaccount (Beller 1990).
Bae nea

40 Chapter Two Matrix Theory in Flux 41

During the formative stage, all of Born’s intellectual pronouncements and permitted freedom from binding epistemological and ontological
were fluid, ambiguous, and uncommitted. assumptions. In the developmentof abstract theoretical physics, it is
2. Born’s initial contribution—hisfirst collision paper (Born 1926c)— tools (mathematical or experimental), rather than preconceived meta-
wasnot written, as Born later claimed, in opposition to Schrédinger’s physical ideas, that constitute the driving force in the growth of knowl-

te
waves. Initially, Born was not involved in any controversy with Schré- edge. And if Born was not committed to particles and indeterminism
dingerat all. In fact, Born was very enthusiastic about Schrédinger’s to begin with, he similarly did not commit himself to wave concepts,
contribution, including its interpretive possibilities. He appears to despite all his enthusiasm for Schrédinger’s theory. Born was instead
have had no strong “belief” in particles: Born was undecided about ’_ working in the creative conceptual twilight where “mathematics knows
the wave-particle issue. better than ourintuition”(interview with Born, AHQP).
3. Born’s aim in hisfirst collision paper (Born 1926c) was not to con- Born’s Collision Papers
tribute to the clarification of interpretive issues but to solve a particu-
lar (yet central) scientific problem. The first paper in which Born provideda statistical description of col-
lision processes using Schrédinger’s formalism is his “Zur Quanten-
4. The aim of the collision papers (Born 1926b, 1926c) was not to argue mechanik der Stossvorgange” (1926c). In his recollections Born implied
the reality of particles and indispensability of indeterminism, as Born’s that his opposition to Schrédinger’s realistic wave concepts preceded
later recollections would lead usto believe, but rather to describe and the actual solution ofthe collision problem, yet a careful reading of the
theorétically’ substantiate Bohr’s concept of “quantum jumps” — dis-
paper does not disclose any opposition. The paper glows sillBorn’s
crete energy changes within an atom during collision processes. Most
of the disagreement between Born and Schrédinger, as their corre-
emotional enthusiasm for Schrédinger’s theory. As Born conceded in
spondence-confirms, centered not on the wave-particle dilemmaor in- this paper, he was unsuccessful in his initial attempts to solve the col-
determinism butrather on the existence of these quantum jumps. lision problem within the matrix framework but was finally able to
obtain the solution with the help of Schrédinger’s formalism. This is
5. The probabilistic interpretation of Schrédinger’s wave function— the reason, Born declared, that he regarded Schrédinger’s formalism as
that % givesthe probability of position—developed not from the rela- “the deepest formulation of the quantum laws” (1926c, 52). -
tively obvious suggestion that y describes the motionof free particles The problem of aperiodic collisions required inquiry into the evolu-
but from the pregnant question of how the wave function is to be in-
tion of atomic phenomena.Earlier matrix methods had answered only
terpreted for bound systems. Born’s interpretationof||? as giving the
probabilities of the stationary states of an atom wasa crucial contri-
questions of structure (energy spectra). Moreover, because thereis a
bution, around which the issues of indeterminism and the particle- similarity between the behavior of the atom in collision situations and
kinematic ontology were elaborated and established. during its exposureto light irradiation, Born hopedthat the solution of
the collision problem would pave the way to understanding the inter-
6. Born’s original probabilistic interpretation played a keyrole in the action between matter and radiation as well.
emergence of the new philosophy of physics not because it was “ob- Born’s first collision paper clearly asserts the superiority of Schro-
viously” correct but because of the ambiguities, difficulties, and para-
dinger’s theory over Heisenberg’s for the solution of this crucial physi-
doxesit raised. The solution to these problemsled to the modification
of earlier concepts andto the elaboration of new theoretical andphilo-
cal problem. Heisenberg’s version of quantum mechanics, argued Born,
sophicalideas. describes only one aspect of quantum problems (stationary states) and
says nothing aboutthe occurrenceof transitions. In contrast, the one-
Born’s probabilistic interpretation can be seen as a “conceptin flux” sentence abstract of Born’s paper claims that “quantum mechanics in
process.’ The uninterpreted mathematicaltools left room for ambiguity the Schrédinger form allows one to describe notonly stationary states
but also quantum jumps” (1926c, 52). Born “intentionally avoids” the
31. The term “conceptin flux” was used by Elkana (1970) to denote those vague and
unspecific ideas that only during the process of simultaneous formation of a theory and 32. As Born recalled manyyearslater: “I wanted to use quantum mechanics because
its basic concepts becomescientifically legitimate. Thus, according to Elkana (1970), in that would give the only way of experimenting with these things. The spectroscopic
Helmholtz’s case the concept of energy and the law of conservation of energy evolved methods give only terms, energies, nothing more . . . the direct way of measuring is by
and were clarified as a single process. The dialogical approach of this book indicates that collisions. Even excitation of light in an atom means collision” (interview with Born, 1963,
all concepts at the focus of research arein fact “in flux.” AHOQP).
 42 Chapter Two
Matrix Theory in Flux 43
term “transition probability,” using instead Schré
dinger’s term “am- himself introduced hisfirst wave mechanical paper merely as a formal
plitudes of vibration.” In fact, Born employed
Schrédinger and de treatment, leaving the question of physical interpretation aside (Schré-
Broglie’s wave concepts quite literally. An atom in
stationary state n is dinger 1926a). Similarly, de Broglie left the meaning of his wave con-
a “vibration process” with frequency (1/h)W® sprea
d over the whole cepts vague: “The present theory may be considered a formal scheme
space. An electron moving in a straight line is,
in particular, such a whése physical contentis not yet fully determined, rather than a full-
vibration corresponding to a plane wave. When the atom
andelec- fledged definite doctrine” (quoted in Jammer 1966, 247). It was natural
tron interact, they produce a complex vibration
(verwickelte Schwingung; at a time when the foundations of physics were shifting to introduce
Born 1926c).%
The wave field determines the atomic trans formal symbolic solutions with minimal interpretive content, especially
itions probabilistically, among mathematical physicists. And Born was indeed “a mathematical
and quite independently of the wayin which the wavefield ofthe scat- |
tered electron is interpreted— literally, or probabilistical method man”(interview with Heisenberg, AHQP). Because there was
ly in terms of no elaborate idea of what de Broglie~Schrédinger waves meant, only
particles.* This is the source of Born’s noncomunit
tal reference to a pu- tentative suggestions, Born preferred to conquer additional territory
tativ e corpuscularinterpretation. As Born concededl
ater: “It is true
that by mathematical means rather than indulge in a controversy over un-
I considered the collision of particles with
other particles as a scattering defined issues. As Born revealed many vears later, his solution to the
of waves” (appendix to letter to Einstein, 13 Janua
ry 1929, Born 1971). . collision problem did not depend on an interpretation in terms of par-
There is indeed noindicationin the paperitself of the
staunch béliefin - ticles: “I cannotseeatall that these purely mathematical objections have
particles that Born professed to have. Rather, Born’s mind
was open on anything to do with the question of particles-waves. . . . For if we ac-
the wave-particle issue.
There are several reasons for Born’s indecision.
oo cept Schrédinger’s standpointthat thereare no particles, only wavelets,
It was.a common the scattering calculations would bé exactly the same as before” (Born
practice at the time to introduce theoretical soluti
ons merely as formal 1953b, 148).
schemes, to be filled with physical content in due time.
Schrédinger Whywas Born openatthe time to Schrédinger’s ideas? Despite the
33. Heisenberg was unhappyaboutthis literal use
of wave concepts: “Ein.Satz erin- fact that Born recalls James Franck’s experiments as a sourceofhis belief
nerte mich lebhaft an ein Kapital aus dem Christl[ichen] in particles, he seemedto consider the wave nature of matter quite se-
GlaubensbeKenntis: ‘Ein Electron
ist eine ebene Welle’ . .. Aber ich will ihnen in
Lastern keine Konkurenz machen” (A riously, as is clear from letter to Einstein (Born to Einstein, June 1924,
sentence vividly reminded me ofan article of faith
from the Christian confession: “An Born 1971).
electron is a plane wave.” ... But Iam not intending
t
to compete with you in slander) De Broglie’s original insights acquired credibility in G6ttingen early
(Heisenberg to Pauli, July 1926, PC).
:
34. To the unperturbed atom with discrete energie
s Born ascribed eigenfunctions in 1925 throughEinstein’s paper “Quantum Theory of the Monoatomic
WE (9x), 08 (qx), . . . ; to the unperturbed electron whose
direction is determined by (a, B, Ideal Gas’ (1924). Justas in his previous treatment of blackbody radia-
y) there correspondeigenfunctions of the type sin 27/A(ax
+ By + yz +). In casethe tion (Einstein 1905), Einstein obtained two terms: one that seemed to
interaction V(x, y, z; q,) is taking place, the scatter
ed wave at infinity will be expressed correspondto a fluctuation dueto particles and another that seemed
throug h (Born does not provide theactual
calculationsin this paper):
. due to waveinterference. In supportof de Broglie’s concept, Einstein
U2 (XY, Z gx) = » if Aa®,.m (, B, Y)
SI Kym (aX + By + yz + 5) B& (4x). concluded that “it appears that an undulatory field is connected with
ext By +tyz>0
where7 is the energy of the incomingelec every motion [Bewegungsvorgang], just as the optical undulatoryfield
tron, coming from the +z-direction, and
is a function of the energy V(x, y, 2; qx)
of the interaction between the atom and
®,,,,. is connected with the motion of light quanta. This undulatory field,
the
electron. whose physical natureis for the momentstill unclear, must in prin-
Concerning the possibility of interpreting this result
in terms of partic
les rather than ciple permitits existence to be demonstrated by the corresponding phe-
waves, Born said simply: “T£ one translatesthis
pie amg
result into terms of particles, only one
is possible: ®,,_,, (@, B, y) gives the probability
nomenaofits motion [Bewegungserscheinungen]” (Einstein 1924, quoted
for the electron, arriving in Stachel 1986, 368). "
rom the z-direction, to be thrown into the direct
ion designated by the anglesa, B, andy
with a phase change 5. Hereits energy t has increa
sed by one quantum hv®,, at the costof We knowthatthese “brief, yet infinitely farseeing” remarks stimu-
the energy ofthe atom” (1926c, 54). lated Schrédinger to look for a wave equation that would describe the
oe function © determines the quant
um transitions n — m, albeit only prob- undulatoryfield of matter. Born,after reading Einstein’s paper, became
abi istically. One doesnotobtain the actual state
of the atom after the collision, only the convinced that “a wave theory of matter can be of great importance”
probability of a certain event (quantum jump).
(Born to Einstein, June 1924, Born 1971), and he supported the efforts
 44 Chapter Two
Matrix Theory in Flux 45
of Walter Elsasser to attribute the Ramsauer effect to the diffraction
As wasfirst pointed out by Linda Wessels (1981), Born did not ini-
of electrons.* It is not surprising, therefore, that when Schrédinger’s
theory appeared, Born immediately recognized its importance and ap- tially connect the wave function with the probability of position; the y-
function controlled the energetic transitions of an atom and the energy
plied it to the problem of collision phenomena, associating a wave with
and direction of motion of colliding electrons. Doubting the possibil-
an electron in the spirit of de Broglie and Einstein.

Ome
ity of ascribing position to intraatomic electrons, Born questioned the
Noris it superfluous to add here that Jordan, Born’s close collabora-
_- tor, also conceived of Born’sinitial collision treatment as being a reality of individual electrons within the atom as well: “Matter can al-
direct ways be visualized as consisting of point masses (electrons, protons),
‘continuation of the idea of matter waves in de Broglie’s, Schrédinger
’s, but in many cases the particles are not to be identified as individuals,
. and Elsasser’s works (Jordan 1927a). What these matter waves exactly
e.g. when these form an atomic system” (1927b, 9). Even though the
meant was as unclear to Born as to Einstein, Schrédinger, and other
presence of particles seemed to be implied by the fact that a “distur-
participants. Born’s elaboration of the corpuscular interpretation, and
bance is propagated along a path away from the atom, and with finite
his later preference for it, as well as Schrédinger’s gradual polarizati
on velocity, just as if a particle were being thrown out,” their existence
to an overall wave ontology, crystallized through the subsequent dia-
should not “be taken too literally” (Born 1927b, 10). (Later, of course, ~
logue amongall the physicists involved.
Born would changehisattitude: ” ||? denotes [the] probability that the
_ It is not clear whether Born wasawareat the time he wrotehis first
electron will be found in the volume-elementdv; this holds in spite of
collision paper that Schrédinger and Pauli had each proved the equiva-
the fact that the experiment, if carried out, would destroy the connec-
lence of thatrix and wave mechanics (Schrédinger 1926f; Pauli
to Jor- tion with the atom altogether”; Born 1969, 147.) :
dan,12 April 1926, PC). But evenif he had been (which is likely), Born Onlyin his second collision paper (Born 1926b) do wefind the begin-
_ could still have considered Schrédinger’s version superior. For Schré-
ning of Born’s opposition to Schrédinger. This opposition was not yet
dinger proved the equivalence between the two theories for bound
militant, though it later became so. Although Born foundSchrédinger’s
systems only. Schrédinger also believed that extensions of the theory
idea of wave packets unsatisfactory and proposed hisown interpreta-
would be more amenable to his version, provingits superiority. Indeed,
tion of the ¢-function as guiding free particles in a “ghostlike spirit”
he suggested the collision problem as a possible case in point (Schré-
in analogy with Einstein’s idea, he nevertheless expressed satisfaction
dinger 1926f). Born’s statementin his paper clearly echoes Schréding-
at the retention of the usual concepts of space-time in Schrédinger’s ~
er’s hopes: “Of the different forms of the theory only Schrédinger’s has
theory. Born did not yet foresee the applicability of his conception of
proved suitable for this process[collisions] . . . and exactly for this rea-
the f-function to the possible motionsof particlesin the interior of the
son I might regard it as the deepest formulation of the quantum laws”
atom.” In his “adiabatic” paper (Born 1927a), Born still did not regard
(Born 1926c, 52).
the 3N-dimensionality of Schrédinger’s wavesfor N particles—later to
Born’s initial enthusiasm for Schrédinger’s ideas is confirmed by the
be one of the major arguments against Schrédinger’s interpretation —
correspondence between the twoat the time. Born admittedin a letter
as a reason to object to a wave ontology.
: to Schrédinger that he was so enraptured by Schrédinger’s workthat, Only with the development of transformation theory (which rein-
zee a5 with “flying banners,” he was drawn back again to the clear conceptual forced and axiomatized Born’sstatistical approach) did Born reach
oes structures of classical physics (Born to Schrédinger, 6 November 1926, his final stand: ““Schrédinger’s achievements reduces[sic] to something
: AHOP). Born was so impressed that he originally considered Schré-. purely mathematical” (Born to Einstein 1926, quoted in Pais 1982,
j dinger’s theory to have more physical meaning than matrix mechan-
f
, terete ro May 1927, AHOQP). This enthusiasm was treatise. Heisenberg from the beginning did not share my opinion that your wave me-
gly expressed, according to Born, in his own paper.°6 chanics is more physically meaningful than our quantum mechanics;yet the treatmentof
~ 8 Ger neat ; ; / the simple phenomenaofaperiodic processes(collisions) led me initially to believe in the
a amsauereffect is a phenomenonin whichthe scattering of electrons during,
superiority of your point of view. In the meantime, I found myself again in agreement
collisions with atomsof certain gases deviates strikingly from the classical theory (slow-
with Heisenberg’s position’ (Born to Schrédinger, 6 November 1926, AHQP).
moving electrons are deflected from their paths much less often than faster moving ones,
37. “The motion of the particles follows laws of probability, but the probabilityitself
= the angle of deflection 1s very small). Elsasser (1925) explained this experimental
: propagates in harmonywith the causallaw.If three stages of the developmentof the quan-
effect as a diffraction phenomenon of de Broglie waves.
tum theory are reviewed, it can be seen that the earliest, that of the periodic processes, is
36. Born wrote to Schrédinger: “You know that immediately after the appearance
of quite unsuitable for testing such an idea. . .. Nothing fundamentalin favor of our
yourfirst works I expressed very strongly my enthusiasm for your conception state-
s inmy — ment can be obtained as long as we consider periodic processes” (Born 1926b, 208).
 46 Chapter Two
Matrix Theory in Flux 47
1198).%* Together with Heisenberg and Jordan, and in contra
st to his structure of energy, namelythecritical potentials that were first
early enthusiasm, Born came to see Schrédinger’s wave mechan ob-
ics as served by Franck and Hertz” (1927b, 11). Born (1927a) further
no more than a mathematical appendix to matrix mechanics. sub-
To Schré- stantiated Bohr’s concepts of stationarystates and discrete
dinger’s total dismay, and despite his desperate pleas for Born to remain transitions
between them bydiscussing the adiabatic Principle in quantu
open to different interpretive options (Schrédinger to Born, m me-
2 Novem- chanical terms.
ber 1926, AHOP), Born allied himself with Schrédinger’
s adamant op- In opposition to Schrédinger, who wanted to do away
ponents. Born’s opposition to Schrédinger’s ideas becam with Bohr’s
e so complete “monstrous” concepts by suggesting an alternative,
that he indoctrinated youngerphysicists against them in continuous wave
a “military,” ontology, Born fully defended’ Bohr’s ideas. Born’s
albeit “playful” manner. Bohr, however, did not agree with (1927a, 1927b) cen-
the Gét- tral inquiry was: to whatextentis it possible to harmonize
tingen attitude, which depreciated entirely the physic the concept
al meaning of of quantum jumps, so fruitful from an experimental point of view,
Schrédinger’s theory, and proposed his complementar
ity as a compro- with
Schrodinger’s wave mechanics, so mathematically powerful?
mise (see chapter 6). As the
Born-Schrédinger andJordan-Schrédinger correspo
ndences reveal, it is
Born’s Probabilistic Interpretation and Quantum Jumps around quantum jumps, and not around waves versus particle
s, that
heated controversy evolved. Schrédinger wrote ina letter to Born that
According to Born’s recollections, every collision experi
ment .con- - he did not rule out the possibility that Born’s ideas were correct. Yet
ducted by his colleague Franck in the adjacent building at Géttingen Schrédinger felt strongly that Boin and‘his colleagues were too ad-
was a clear demonstration of the corpuscular nature of electrons, thus dicted to the oldconcepts of stationary states and quantum jumps. He
prompting Born to suggest an interpretation in direct opposition to could not comprehend whyaninterpretation that does not dogmati-
Schrédinger’s ideas (Born 1961; interview with Born, AHOQP). I have cally postulate the discontinuities, as the G6ttingen-Copenhagen does,
already argued thatthis story does not withstand histori
cal scrutiny. is a priori impossible. His own continuoustreatment of resonance
Franck’s name is indeed essentialin anotherrespect: his famous be-
work tween two atoms (Schrédinger 1927b) led to the same consequences
as
(with Gustav Hertz) on atom-electron collisions was generally
regarded the theory that postulated discontinuous exchanges of energy (Hei-
as direct experimental evidence for Bohr’s ideas about stationary
states senberg 1927a). Does notthis indicate, inquired Schrédinger, that one
and the discontinuous energetic transitions between them— quantu should explore whetherall other discontinuities can be similarly de-
m
jumps.In fact, the theoretical quantum mechanical description of these duced rather than dogmatically postulated (Schrédinger to Born, 2
Jumps was Born’s central aim,as is clear from hisfirst and second col- No-
vember 1926, AHOP)?
lision papers. The abstract ofthefirst collision paperclearly states
its In his reply to Schrédinger, Born conceded thatafter the appearance
main purpose: the theoretical description of quantum jumps with
the of Schrédinger’s theory, he was atfirst disposed to return to the clear
help of Schrédinger’s formalism. Born summarized the
essence of his conceptual frameworkof classical physics. But he changed his mind. Of
Es collision papers as a “precise interpretation of just these observa
tions course, Born agreed with Schrédinger,it is desirable to explore all pos-
which may be regarded as the most immediate proof of
sa

the quantized sibilities, yet he himself did not intend to do so. Instead, Born
had de-
cided to rely on his ownfeeling that one cannotdispense with quantum
38. Born was impressed byPauli’s argumentthat regular position
and that by a suitable canonical transformation the descript
space is not essential, jumps (Born to Schrédinger, 6 November 1926, AHPQ).
ion in position space can be In his secondcollision paper (Born 1926b), Born spelled out the con-
converte
cet

d into a description in momentum space (see

Pauli to Heisenbe rg, 19 October
1926, PC; Born to Schrédinger, 6 November1
926, AHQP).
nection between a system’s stationary states and Schrédinger’s wave
39. According to Jordan: “Finmal in G6éttingen bei function. For a periodic system (an unperturbed atom, say) represen
einem Spaziergang von Born mit ted
Dirac und Oppenheimer und vielleicht mit noch
einem oder zwei anderen . . . hat Born by W(x) = =C,w,(q) (discrete spectra), Born proposed that |C,,|? is the
also einen Brief von Schrédinger vorgelesen und Oppenhe
Born hatte so wie ein Feldherr seine Truppen
imer erzahlte mir scherzhaft, “frequency”of the state n in a group ofidentical noncoupled atoms. In
zusammenruft, so hatte er den anderen other words, |C,,|? denotes the probability that the system is in a sta-
erklart, was fiir falsche Ideen das waren die
Schrédinger da hatte” (Once on a walk in
Géttingen with Dirac and Oppenheimer andperhaps
with one or two persons... Born tionary state described by ,(q).
read a letter from
Schrédinger aloud, and Oppenheimer told
me jokingly that Born,as if Born’s interpretation of |C,,|? as giving the probability of the station-
he were the commanderin chief summoninghis
army, explained to everybody howfalse ary states of the system was a pregnant idea. It shifted the probability
were Schrédinger’s ideas at the time) (interview
with Jordan, AHQP). considerations one level deeper, from transitions to stationarystates,
 48 Chapter Two
Matrix Theory in Flux 49

thus making the transition probabilities calculable from the moreele- Pauli, who had stood somewhataloof from the developmentof ma-
mentary probabilities of these states. This point would be crucialin the trix mechanics, was not as hostile to Schrédingeras Heisenberg was,
Copenhagen deliberations on indeterminism (see chapter3). nor did Pauli subscribe to the opinion that the positions of intraatomic
Born and his contemporaries initially viewed the probabilistic inter- electrons are unobservable in principle.” Thus Pauli favored resuscitat-
pretation of stationary states as the crux of his interpretive achieve- ing’the visualization of the stationary states throughthe kinematic data
ment.” Precisely because this hypothesis for atomic states and transi- of positions of intraatomic particles. Pauli’s contribution, which stimu-
tions was farther reaching and morespeculative than Born’srelatively lated Heisenberg’s own revival of regular space-time through the un-
straightforward, even obvious, suggestion for free particles, he and his certainty principle (see chapter 4), was indeed a landmark on the way
_ contemporaries saw his interpretation of |C,|? as a possibly fruitful to the establishmentof a particle ontology of quantum physics. No such
conjecture, still in need of experimentalverification, rather than an “ob- particle ontology had underlain Born’soriginal interpretation.
viously” correct interpretation."! Born’s owncalculations for the collisions between electrons and hy-
It was Pauli who generalized Born’s suggestion concerning the prob- drogen atom were carried out only to a first approximation. Therefore,
ability of stationary states and endowed this concept with particle- Born concluded, “it would be decisive for the theory if it should prove
kinematic meaning. Born deemed it meaningful only to talk about possible to carry the approximation further” (1927b, 13). In his text-
energies and angular momentaofstationary states. Pauli, however, de- book Atomic Physics, written in 1935, Born emphasized the importance
ae fined lq. ---,4p)|2dqy - ‘dq, as the probability that, for a definite sta- of verifying his original hypothesis by subjecting scattering and other
tionary state, the coordinates q, of the particles (k= 1,..., f) lie between phenomenato precise quantum mechanicalcalculations (Born 1969).
qx and q, + dq, (Pauli to Heisenberg, 19 October 1926, PC). This was Born mentioned in this respect his and Vladimir Fock’s work on the . -
actually the first time the wave function had been described as giving excitation of atomic systems, Dirac’s work on the excitation of coupled
the probability of the positions of intraatomicparticles. Pauli’s defini- systems, Born’s and Mott and Massey’s work on collisions, and Went-
tion later appearedin print, in a footnote to a paper on gas degeneracy zel's verification of the Rutherford formula. Such experimental con-
and paramagnetism (Pauli 1927, 83). firmation was, according to Born, more important than philosophical
elucidation. © “wee wl. &
40. Jordan, discussing recent developments in physics, characterized Born’s contribu- This process of critically evaluating and modifying Born’s original~
tion as allowing the determinationof the probability that an atom is in a definite station-
ideasis still going on. Eugen Merzbacher(1983) has described advances
arystate at a definite moment in time: :
In Verfolgung seiner Vorstellungen hat Born die physikalische Bedeutung der in ion-atom collisions as, in many cases, substantiating Born’s work.
Schrédingerschen Wellenfuktion scharfer bestimmen kénnen. Er hat gezeigt, dass AbrahamPais (1982) has described his own success in carrying out a
man mit ihrer Hilfe ein Maass der Wahrscheinlichkeit dafiir angeben kann, dass calculation for scattering by a static, spherically symmetrical potential,
ein Atom zu einem gewissen Zeitpunkt sich gerade im n-ten Quantenzustande andhisfailure to extend Born’s methodtorelativistic field theories. “To
befindet. Diese Deutung der Schrodingerfunktion konnte gestiitzt werden durch this day, proof or disproof of the convergenceof the Born expansion in
den Nachweis, dass das bekannte Ehrenfestsche Adiabatenprinzip der Quanten-
iat theorie im Anschluss an diese Deutung auch vom Standpunkte der neuen Quan- field theory remains an important challenge yet to be met” (Pais 1982,
1198). Recent work on scattering deals with beth mathematical im-
Se
tenmechanikformuliert und bewiesen werden konnte. (1927a, 646)
(In further developing his conceptions, Born succeeded in defining more provements and experimental physicaldifficulties (Daumer 1996; Cush-
precisely the physical meaning of Schrédinger’s wave function. He demon- ing 1990). ;
strated that with its [wave function’s] aid, one could determinethe value of
- the probability for an atom to bein the nth quantum [stationary] state at a
The meaning and the foundational status of the Born-Pauli probabi-
certain momentin time. This interpretation of Schrédinger’s wave function listic interpretation remain today a subjectoflively controversy (Durr
was corroborated by demonstrating that Ehrenfest’s familiar Adiabatic et al. 1992a, 1992b, 1996; Valentini 1996). What appearsin textbooks and
Principle of quantum theory could also be formulated and proved from the philosophical writings to be a closed box has never beensealedoff at
standpoint of the new quantum mechanics in conformity with this [Born’s]
the frontier of research.
interpretation.)
41. Born emphasized this pointat the time: “It is, of course, still an open question
whether these conceptions can in all cases be preserved. . . . Unfortunately, the present 42. According to Leon Rosenfeld (1971), Pauli did not welcome a complete break be-
state of quantum mechanics only allows a qualitative description of these phenomena tween the formalized conceptual scheme of matrix mechanics andtheclassical notions of
{collisions of electrons with atoms—in particular, collisions of electrons in helium]’”’ space-time and Keplerian orbits that underlayit.
(1927b, 11-12).
 CHAPTER 3
5
»
Quantum Philosophyin Flux

- The external conditions; which are set... by the facts ofexper

ience, do not
permit him [the scientist] to let himself be too muchre
stricted in the construc-
‘tion of his conceptual world by the adherence to an epistem
ological system.
Hetherefore must appear to the systematic epistemologist as a
type of
: unscrupulous opportunist.
Albert Einstein 1949b, 684

Introduction

’. In the previous chapter I argued against the widely accepted miscon

-
ception that the G6ttingen-Copenhagen physicists developed matrix
mechanics by implementing a committed indeterministic corpuscular
ontology as opposed to Schrédinger’s causal wave ontology. I also ar-
gued that the two approaches were not conceptually distinct and his-
torically independent. I described the development ofthe G6ottin
gen-
Copenhagen version of quantum mechanicsascharacterized by vacil-
lation on foundational questions and opennessto the wave theoret
ical
perspective. Perhaps the strongest expression ofthis openness, if not
Preference, is a statement Max Born madeduring his lectures in the
United States in 1926: Heisenberg’s matrix elements are not amplitudes
of radiation but “real waves of an atom” (1926a, 70). The polariz
ation
and conflict between the wave theoretical and matrix approaches was
indeed the result of the threat Schrédinger posed to matrix theorist
s;
acknowledgments of mutual debt were largely eliminated from the
public discourse. Yet the developmentof the Gottingen-Copenhagen
version depended, in part, on the de Broglie—Schrédinger
ideas—as a
stimulus to response and as a resource for selective appropriation.
De
Broglie’s work reinforcedBohr’s realization that one had to dispen
se
with particle space-time models in the quantum domain (Bohr 1925);
this realization was implemented in Heisenberg’s reinterpretation
paper
(1925). Wave theoretical imagery informed Jordan’s early treatment of
electrodynamics in matrix mechanics and guided his pioneering work
on the theory of quantized matter waves (Darrigol 1986; Kojevn
ikov
1987).
 52 Chapter Three Quantum Philosophy in Flux 53

Another widely held misconception is that a positivist philosophy the atom,” Born claimed, for the same epistemological reasons that
of “elimination of unobservables” was central to the emergence of the Einstein had eliminated the concepts of absolute velocity and absolute
theoretical structure of the new quantum theory. It was this positivist simultaneity.
philosophy, so the story goes, that guided Heisenberg’s efforts from his Historical research leads to a different picture (MacKinnon 1977;
reinterpretation paper to his uncertainty paper (Hendry 1984). Mehra and Rechenberg 1982; Beller 1983, 1988; Hendry 1984). Dur.
I argue that positivist philosophy wasless a heuristic principle and ing the years preceding Heisenberg’s reinterpretation paper, quantum
more a tool with which theoretical advances could bejustified ex post physicists increasingly encountered the inadequacyofthe notion of in-
facto. Contrary to the received opinion that Heisenberg’s philosophical. traatomic orbit. When physicists questioned the adequacy oforbital
stand remained stable from the reinterpretation (1925) to the uncer- notions, their doubts had more to do with the theoretical failure of or-.
tainty paper (1927b), a careful analysis reveals a radical change, if not bits than with their experimental unobservability. Orbital assumptions
an about-face. Heisenberg employed two different strategies of justifi- failed in the domain of the interaction of light with matter; they could
cation: the Macheanpositivist principle of elimination of unobservables notbereconciled with the fact that the dispersion oflight occurs with
in his reinterpretation paper and the operational approach ofdefining spectroscopic rather than orbital mechanical frequencies. Moreover, in
physical concepts through the procedure of their measurement in his the domain of the constitution of atoms and molecules (with the full
uncertainty paper. If in 1925 Heisenberg claimed that directly observ- price of ad hoc assumptions), all modificationsof the orbital mechanical _
. able experimental data (frequencies and intensities of emitted radia- models failed to do justice to the experimentalstate of affairs (Hendry-
tion) should determine theoretical structure, in 1927 he declared that 1984). ,
it is theory that gives meaning to experiment: “It is the theory which Nordid it help much to attach to these models a formal symbolic,
decides what we can observe” (1971, 77). No coherent philosophical rather than realistic, significance. The orbital model failed for the
choice between positivism and realism guided Heisenberg’s efforts. A anomalous Zeeman effect, and for the spectrum of the helium atom. As
fascinating, ever changing mixture of realist intuition and positivist Born attempted to argue in 1924, alt possibleorbital assumptions led to
legitimation characterizes Heisenberg’s work leading to, and springing equally wrong results (Hendry 1984).? Gradually it becameclear that
from,the reinterpretation paper. there wasreally no good reasonto cling to the failed orbital concepts
and the mechanical models built on them. The need nowarose, as Pauli
Positivism in Flux understood, for a new, more adequate kinematic law of motion and
perhaps too for anew dynamic law responsible for the changesin kine-
The myth of the fundamental heuristic role of the elimination of unob- matics. It seemed reasonable at this point to introduce the so-called vir-
servables originated with the matrix physicists themselves. Born de- tualoscillator model, first applied to the interactionof light and matter,
clared in his presentation of the new theory in winter 1925-26 (as well and soon extended to theories of atomic structure. In the domain ofthe
as in later writings) that Heisenberg endedthecrisis in physics by intro- interaction of matter with radiation, this symbolic model proved to be
ducing a fundamental epistemological principle that could be traced a powerfultool for translating and extending successfulclassical for:
back to Emst Mach. This positivist principle states that concepts and mulasinto the quantum domainvia the correspondenceprinciple (Dar-
representations that do not directly correspond to observable “facts” rigol 1992a). Reaching the conclusionthat “reality was notin orbits, but
are not to be used in a theoretical description of reality (Born 1926a, ratherin Fourier components, or rather their quantum mechanical ana-
68—69).! Heisenberg, supposedly, was encouraged by Einstein’s ex- logues” (interview with Heisenberg, AHQP), and encouraged byhis -
ample—in 1905 Einstein ended another majorcrisis in physics using previous results (Kramers and Heisenberg 1925), Heisenberg decided
the principle of elimination of unobservables. Heisenberg banished the in the middle of May 1925 to attack the problem of hydrogenlineinten-
picture of electron orbits and the conceptof “electron position within sities. Heisenberg’s correspondence with Ralph Kronig, whom hekept
informed abouthis progress, clearly reveals that epistemological con-
1. Instead of such unobservable kinematic variables as the position, velocity, or pe- siderations were far from Heisenberg’s mind during hisfirst attempts
riod of revolution of an electron in an atom, Heisenberg incorporated experimentally
observable spectroscopic data (frequencies and intensities of radiation) into the theoreti- 2. Today we know thatat the price of nonlocality one can retain space-time trajectories
cal framework. within an atom (for example, Bohm’s theory).
/
 54 Chapter Three
Quantum Philosophy in Flux 55
to tackle the problem.? The hydrogen problem was too complicated
adhere to introducing only those quantities in physics which
mathematically, and Heisenberg hadto content himself with the simple are ob-
servable in principle” (quoted in Hendry 1984, 19).
case of a harmonic oscillator. Even in this case, mathematical difficul-
However, the key phrase hereis “in principle.” The question
ties remained, and Heisenberg wasnotyet sure what physical conse- was not
whether the force acting on the electron could be experi
quences his new schemeimplied: “The physical interpretation of the mentally mea-
sured but whether the definition of the conceptoffield strength
above mentioned schemeyields very strange points of view” (Heisen- was
consistent with the hypothetical possibility of measurement.
berg to Kronig, 5 June 1925, AHOP). Heisenberg confessed to Pauli Because
that the electron was assumed to be the smallest
his private philosophy was simply “a mixtureofall possible moral and test body, it was meaning-
less (according to the operational approach)to talk about
aesthetic calculations and rules” through which he did not “findhis space-time in
the electron’s interior, but it was not meaningless to
way anymore” (Heisenberg to Pauli, 24 November 1925, PC). Ina letter talk about the po-'
sition of this smallest test body in the larger atom. There was nothin
to Pauli, Heisenberg arguedthat the interpretation of the experimental g
inconsistent in assuming that in principle electron position
formula for the hydrogen spectrum onthebasis of Kinematic orbital could be
measured inthe interior of an atom, though thestate of experim
concepts was impossible: “An interpretation of the Rydberg formula in ental
techniques did not allowit in practice. One could, for example
terms of circular andelliptical orbits does not have the slightest physi- , devise
a y-ray thought experiment—such as Heisenberg proposed later in his
cal significance.” He addedthat his efforts were devoted to “killing the uncertainty paper. And if electron positions are not unobservable
concept of an orbit which cannotbe observed anyway” (Heisenberg to, in
principle, the same holdsfor electron orbits. Pauli’s stand, pushed
Pauli, 9 June 1925, PC). The order of precedence seemsclear: orbits are to its
logical conclusion, implied merely that intraatomic positions and
theoretically inadequate, and they had therefore better be eliminated. trajec-
tories could be determined only with the accuracy of the size of the
One can dispense with orbits without regret because they do not have electron—the smallest particle knownatthe time. And,in fact,
any observational significance. Pauli
. did emphasize in a letter to Eddington of 20 September 1923 (PC)
A comment on the possible influence of Pauli on Heisenberg is in that
the electron’s position need not be considered unobservable in prin-
place here. Most historians whotreatthis subject (Hendry 1984; Serwer ciple, and that the question of technical difficulties should notenter into
1977) assume that Pauli’s operational attitudecalled for the elimination considerations of principle connected with the definition of physica
of the orbits and positions of intraatomic electrons because they are, as l
concepts.4
Hendry puts it, “operationally meaningless.” Heisenberg’s reinterpre- As already mentioned, Heisenberg was led to his reinterpretation
tation paper is often perceived as an implementation of Pauli’s episte- procedureby trying to solve the problem of hydrogenintensities. His
mological program. Yet this point of view needsto be qualified. attempt did not succeed. Heisenberg wasforced, bytechnical difficul-
It is true that, at least after 1919, Pauli advocated in certain theoreti- ties, to stop at the programmatic point. Had he solved this problem,
cal contexts an operationalattitude toward physics. For example, he Heisenberg’s motto “success sanctifies the means” would suffice to
claimed that the strength ofelectric fields in the interior of an electron jus-
tify the procedure of replacing the classical coordinates with a set of
isa meaningless concept because thefield strength is defined as.a force quantum theoretical magnitudes. Yet at this programmatic point, Hei-
acting on a test body, and’ there are no smaller test bodies than the senberg needed a more general conceptual justification, and he chose
electron itself. Similarly, the concept of continuous space-time in the the principle of elimination of unobservables. This principle
interior of the electron is meaningless. Pauli argued that “one should , sup-

4. Not by accident did Pauli fail to mention the principleof eliminatio

3. Ina letter to Kronig, Heisenberg’s program is clearly stated:In classical theory, n of unobserv-
Hei- ables when presenting the essentials of Heisenberg’s approach in his
senberg wrote, knowledge of the Fourierseriesis sufficientto calculate the radiation com- paper on the hydro-
gen atom (Pauli 1926b). Significantly, it was Pauli himself wholater restored
pletely, not only dipole moments but higher momentsas well. In principle, one should the kinematic
be positions of intraatomic electrons in his extension of Born’s
probabilistic interpretation.
able, by replacing classical variables with their quantum analogues in the equation
of 5. Why did Heisenberg not use Bohr’s correspondence principle instead,
motion, to obtain an exact, complete expression for the intensities. Because a principle
Heisenberg that had played a most fundamentalrole in his reinterpretation procedure?
wanted to take into consideration quadruple and higher moments, he had to immerse The most
probable reasonis the collapse of the Bohr-Kramers-Slater theory (1924)
himself in the technical mathematical problem of what multiplying X and Y meant, and its associ-
when ated research program.In this theory, the correspondenceprinciple, the
Xand Y were algebraic sets of quantum substitutesfor classical
position coordinates (Hei- ~
wave nature of
radiation, and nonconservation were tightly connected—andtherefutatio
senberg to Kronig, 5 June 1925, AHQP). n of the Bohr-
Kramers-Slater theory by the Bothe-Geiger experiments seemed to indicate
that the cor-
56 Chapter Three
Quantum Philosophy in Flux 57

ported by Einstein’s authority, was a clever strategic choice. Precisely atom and theradiation it emits. Heisenberg’s “realistic” presupposi-
because it was only invoked a posteriori it played no essential role in tion “that something in the atom must vibrate with the right fre-
the derivation of the formulas in the reinterpretation paper (1925). The quency” (Heisenberg to van der Waerden, 8 October 1963, van der
reinterpreted formulas held true whetheroneheld the position of elec- Waerden 1967) prompted him to construct a bizarre noncontinuous
trons to be observable or not. The matrix formalism was not under- space (througha reinterpretation of kinematics) so that electron “mo-
mined after Heisenberg changed his opinion aboutthe observability of tions” in this space would have the correct spectral frequencies. The
electrons in his uncertainty paper (1927b). The above analysis demon- advantage of gaining a physical sense of the underlying atomic model
strates that the elimination of unobservables was,in fact, not a guid- came at the price of abandoning visualization in regular space-time
ing principle, but rather a generaljustification of a powerfultechni- and wasjustified by the positivist ideology of elimination of unob-
cal method that de facto eliminatedclassical positions and orbits. The servables.’
elimination of the space-time container and the loss of visualization Schrédinger shared Heisenberg’s conviction that a connection must’
were prices to be paid, not goals to be attained. Not surprisingly, the exist betweenelectron motion and radiation frequencies, yet he did not
authors of matrix mechanics sometimes would talk not in termsof “ob- have Heisenberg’s inclination to operate in nonclassical spaces torn, as.
servable in principle” but rather in terms of what “can be observed ex- Schrédingercalled it, by “yawns” and “gaps.” Schrédinger announced
perimentally” or whatis “really observable” (Born 1926a,69).° that “what we cannot comprehend within it {regular space-time] we --
One of the weapons Géttingen-Copenhagen physicists wielded cannot understandatall” (1926b, 27). He offered his own wave mecha-
against Schrodinger’s competitive approach was an accusationof anach- nism, which preserved the direct relationship between electron mo-
ronistic, “naive” realism. Yet Heisenberg and Schrédinger, despite a tions and radiation frequencies. The contrast between Heisenberg and
difference in intellectual style, were not far apart on the issue of real- Schrédinger is not adequately described as a realist-positivist dichot-
ism. Both employed visualizable intuitions; both willingly sought new omy. The difference between them lies elsewhere. | ,
“artifices of thought” when old onesfailed; both engaged in heated Shifting theoretical and social circumstances forced Heisenberg to
controversies about the “reality” of certain scientific notions (quantum abandonthepositivist position that he advocatedin his reinterpretation
discontinuities, ~-functions). In fact, heated controversies about unob- paper. In his uncertainty paper, Heisenberg emphasized not the close
servable “reality” played a decisive part in the genesis of interpreta- proximity of basic theoretical terms to direct sense perceptions (what
tions of quantum physics. Thusthe reality of quantum jumps was the can be closer to Machean sense perceptionsthanthe frequencyandin-
central issue in the interpretive efforts undertaken duringthe crucial tensity oflight, or, in other words, light’s color and brightness?) butthe
years 1925-27. Indeed, this issue channeledtheefforts of the “positiv- supremacy of theory over facts (a good theoryis notrefutable by facts;
ist” group made up of Born, Heisenberg, and Jordan nolessthan it did a good theory is needed to define whatthefacts are).
those of the “realistically” inclined Schrédinger (see chapters 4 and 6). Heisenberg’s interpretive attempts can hardly be viewed as a smooth
When,in his reinterpretation paper, Heisenberg replaced the familiar evolutionary process, guided by consistent epistemological concerns.
classical parameters of motion with abstract algebraic constructs, the What we have witnessed, rather, are changes of opinion onbasic is-
basic physical intuition behind his innovation was that there must be sues, trial and error, about-faces. But though extraordinarily agile, Hei-
a correspondence between the internal dynamical “mechanism” of an senberg wasnot the only physicist whoskillfully adapted himself to
changing theoretical opportunities. Pauli, for example, expressed the_
respondence principle could no longer lead to progress. Indeed, Einstein objected in a opinion in 1925 that probability statements should not enter into the
letter to Heisenberg that his approach in the reinterpretation paper wastoo close to the basic postulates ofa satisfactory physical theory, only to becomeoneof
discredited Bohr-Kramers-Slater theory. Serious doubts aboutthe fundamental role of the
correspondence principle had been expressed even beforethe Bothe-Geiger experiments.
the architects of the probabilistic kinematic interpretation of the wave
Sommerfeld, Heisenberg’s teacher, declared in 1924 that the correspondence principle
was not part of the foundation of quantum theory—it was merely its limiting theorem 7. Heisenberg’s physical intuition was obscured by Born andJordan’schillingly
(Mehra and Rechenberg 1982). Pauli similarly denied the ability of the correspondence elegant
matrix formalism; Heisenberg’s original awkward mathematical formulation contained
principle to be a faithful guide for understandingthestructureof atoms (Mehra and Rech- physical heuristics that did not appear in the matrix formulation. This is exactly
enberg 1982). what
Pauli feared when herefused to collaborate with Born, telling him that his “tedious
6. A more detailed analysis of the role of operationalism in the erection of quantum and
complicated formalism ..is only going to spoil Heisenberg’s physical ideas” (quotedin
theory is given in Beller (1988). ‘ van der Waerden 1967, 37).
 58 Chapter Three
Quantum Philosophyin Flux 59
function two yearslater. Nor do wefind consistency between the epis-
Indeterminism in Flux
temological views of quantum physicists andtheir actualscientific prac-
tice. Pauli, who asserted the needto replace orbits in atoms with more argued in chapter 2 that one should notascribe to Born any commit
satisfactory kinematics two years before Heisenberg’s reinterpretation, ted
opjnions about particle ontology or aboutthe physical significance
continued research based on orbital assumptions. Einstein, at least in of
Schrédinger’s theory; these were issues that Born initially left open. My
his youngeryears, refused to be restrained by an epistemologicalstrait- claim applies similarly to the problem of indeterminism. Heisen
jacket. In his discussion with Pauli in the 1920s, when Pauli argued that berg, I
have argued (chapter 2), was open on the question
the continuum must be abandoned on operational grounds, Einstein - of determinism ver-
sus indeterminism.I will argue, as well, that Born did not have a com-
replied that the issue must be decided only on the basis of what was -. mitted opinion about causality, only tentative, uncommitted
theoretically functional (Hendry1984). sugges-
tions.? My analysis of Born’s stand will focus on two issues: the nature
Nor did Niels Bohr consider philosophical analysis heuristically valu-. of positivist statements in quantum mechanics and the problem ofele-
able. He did not take seriously considerations of simplicity, elegance, or
mentary probabilities. :
even consistency (the “epistemic virtues”), holding that “such quali- Born’s indeterminism in his first collision paper wasrather hesitant:
ties can only be properly judged after the event” (Rosenfeld 1967, 117). “I myself am inclined to give up determinism in the world of atoms.
A harsh and crisp verdict came from Paul Dirac: “Ifeel that philoso- Butthatis a philosophical question for which physical arguments alone
phy will never lead to important discoveries.It is just a way of taiking are not decisive” (1926c, 54). He wavered even morein his second col-
about discoveries which have already been made” (interview with lision paper(1926b):even though Born himselfinclines toward indeter-
Dirac, AHQP). bas minism, he fully realizes that attempts to look for “hidden” internal
While the above quotes seem to indicate that philosophy is always parameters, which determine individual processes, are consistent with
used only a posteriori, in the “context of justification,” as. opposed to the prevailing state of quantum theory. And evenlater, when Born did
the “context of discovery,” such an assertion would betoo categorical. commit himself publicly to indeterminism, he confessedin letter to
Quite apart from thefact that a meaningfuldistinction between the con- Einstein that the issue is undecidable in principle (Born to Einstein,
text of discovery and the contextofjustification cannot always be sus- 13 January 1928, Born 1971).
tained, the sources of scientific creativity can be foundin diverse fields Born’s main concernat the time was neither deep conceptualanalysis
of human activity— philosophyis not the most unlikely of them. Philo- nor philosophical deliberation, but the solution of the collision prob-
sophical ideas might be suggestive inparticular theoretical settings: lem. As Born indicated in an interview with Kuhn,he did not think his
the idealist German philosopher Fichte might have been a surprising solution involved any far-reaching implicationsfor the issue ofindeter-
source of Heisenberg’s idea of the reduction of a wave packet (see chap- minism: “I did not consider it very philosophical. I thoughtgiving up
ter 4). Historians of science have described extensively the rich cultural the description in space and time and replacing it by symbolic descrip-
and political milieu in which the development of quantum mechanics tion was much deeper and much morephilosophical. Andto finda way
took place (Faye 1991; Forman 1971, 1979; Heilbron 1987; Holton 1970; of expressing it in simple terms—probability——seemed to me not so
Krips 1996; Wise 1987). Yet these contexts served moreas resources and very important... .We were so accustomedto makingstatistical consid-
less as influences. The point is not that philosophy cannot influence erations, andto shift it one layer deeper, seemed to us not so very im-
science. Creative scientists might adopt a certain foundational stand
portant” (interview with Born; AHQP).
(sometimes indistinguishable from a traditional philosophical stand) in The description of collision phenomenathat Born provided was sta-
order to pursue a definite line of research. Such a philosophical orien- tistical: one does not get an answerto the question “Whatis the state
tation is, however, local and provisional. The longevity of philosophical after collision?” but rather to “Howprobable is a specified outcome of
“commitment” is coterminous with its usefulness in solving the prob- the collision?” (1926c, 54). Because the necessity of indeterminism was
lem at hand.®
9. Fora different opinion, see Forman (1979). I use the words “causality” and “deter-
8. A scientist can develop a preferred philosophical position, yet such a position crys-
minism” interchangeably here, following the usage of quantum physicists. In a careful
tallizes only after an entire career, when certain basic characteristics of successful theories
philosophical discussion these notions need to be dissociated. Fora wide-ranging analysis
that the scientist has developed are internalized and “entheorized.” For the notion of of the complexity of the notion of determinism, and of the application of this notion in
“entheorizing,” see Fine (1986) and chapter9.
different branches of modern physics, see Earman (1986).
earmeneand
 60 Chapter Three
Quantum Philosophy in Flux 61

far from obvious, Born had to face a natural question: washis solution Born made, not decisive statements about the necessity of indetermin-
onlytentative and a more complete, deterministic description possible
ism, but rather mild arguments concerning the agreementof theoreti-
in the future, or was his contribution fundamental and not susceptible cal indeterminism withan existing practical statistical approach. Actual
to any substantial modification? In short, Born was looking for argu- experiments with collisions deal de facto only withstatistics: “Ought
ments to legitimate his contribution, as is often done in theoretical we to hopelater to discover such properties (like phases of the inter-
physics.” Moreover, Born had a special need for such justification be- nal motions) and determine them in individual cases? Or ought we to
cause, in his treatment, the statistical conclusions followed from exact believe that the agreement of theory and experimentas to the impossi-
data: even though the momentum andenergyof the incomingelectron, bility of prescribing conditions for a causal evolutionis a preestablished
as well as the energy of the stationary state of the atom, were given, his
harmony foundedon the nonexistence of such conditions?” (Born 1926c,
description of the result of the collision was probabilistic, This situa- 54).% “Preestablished harmony” is not a forceful argument, but Born
tion was clearly unsettling after the Bothe-Geiger experiments, which had no better—yet. The difference between classical and quantum
seemed to indicate that individual atomic phenomenaare determinate probabilities wasyetto be realized, probabilistic transformation theory
(that conservation laws apply strictly; Bothe and Geiger 1925a, 1925b). wasyet to be established, Heisenberg’s uncertainty Paperwas yetto be
Born tackled this problem from two different angles. He denied an written. /
analogy between his solution and the Bohr-Kramers-Slater theory— In a nontechnical paper that summarized Born’s positionatthetime,
discredited by the Bothe-Geiger experiments—despite a very close wefind no strong arguments for indeterminism: “The quantum theory .
resemblance between them (Born 1927b, 10 n; Born to Schrédinger, is in agreement with the experimentalists, for whom microscopic coor-'
6 November 1926, AHOP).1! The conservation laws in collisions are dinates are also out of reach, and whotherefore only countinstances
enforced, Born claimed, by the basic formalism of quantum theory and indulgein statistics. Itis not forbiddento believein the existence of
(1927b, 10 n). Hetried to justify his statistical solution by positivist these coordinates, but they will only be of physical significance when
arguments: because formal quantum mechanics renounced “internal methods have beendevisedfor their experimental observation” (1927b,
atomic motions,” there is “no reason to believe that there are some inner 11). As long as such a possibility was not ruled outin principle, Born
properties of the atom which condition a definite outcomeforthe col- should have remained (and did in fact.remain) undecided on the issue
lision” (1926c, 54). of indeterminism. So
Yet Born must have realized at the time that the renunciation of .- There are additional, theoretically deeper reasonsfor the indecision
space-time as a container of motion was far from necessary. As I have amongphysicists on the issue of indeterminism.It is often supposed
mentioned, Heisenberg’s claims that electron position within an atom in historical accounts that because quantum theory wasdefacto statis-
is in principle unobservable were faulty: intraatomic position is unob- tical, the creators of quantum theory had some prior indeterministic
servable notin principle but merely in practice. And because positivist beliefs, or at least strong predispositions. The typical story runsas fol-
arguments for the nonexistence of internal atomic parameters, or “hid- lows: Bohr talked aboutprobabilities of spontaneous emission in the
den variables,” were inconclusive, similarly inconclusive were asser- 1920s; therefore, he and the Géttingen-Copenhagen groupalready be-
tions about indeterminism based on such arguments. This is the reason lieved in indeterminism at that time. Only the conservative Einstein
(1917) regarded the introduction ofa priori transition probabilities as
10. Discussing quantum mechanical and semiclassical solutions in collision processes,
Merzbacher(1983) remarked: “The virtues or disadvantagesofa partially, or even fully, a weaknessof the theory, because he could not accustom himself to a
classical approach are frequently contrasted with the quantum mechanical approach,but “dice-playing God.” : ee
since convenience often dictates the choice, such comparisons are usually biased by the But this account does not hold up against the historical record. |
desire to justify a particular calculation after its completion.” know of no quantum physicist before 1927 who did commithimselfto
11. The analogy between the Bohr-Kramers-Slater theory and Born’scollision theory is
indeterminism.Physicists at the time thought statistically along classi-
in fact very close. The electronic “phantom” ¢-field probabilistically controls the ener-
getic transitions of atoms in collision, in the same way that the virtual radiation field cal lines (the uniqueness of quantum probabilities was recognized only
probabilistically controls atomic transitions in the Bohr-Kramers-Slater theory. The action
of the d-field is twofold: according to Einstein’s idea of light quanta in the case of free
12. That such a “preestablished harmony” existed between quantum theory and the
particles andin analogyto the Bohr-Kramers-Slater theory in the case of bound particles
possibility of experiment was in fact argued by Heisenberg (1927b) in his uncertainty
(an atom).
paper, though withoutthat particular term being used.
 62 Chapter Three
Quantum Philosophy in Flux 63
after Born’s interpretation). In classical statistical theory one starts
with probabilistic interpretation by Born andits extension by Pauli
an assumption about equally probable cases (elementary probabilities) indicated
the possibility of building a consistent quantum probabil
and derives from them more complicated probabilities (Born 1927b;Jor- istic calculus,
dan 1927b). In quantum theory, however, complex probabilities (such and thus the possibility of reaching a final verdict on indeter
minism:
as transition probabilities) were introduced a priori, without being “Itis thus very significant that in Pauli’s above-mentioned
re- formulation
nothing is said about a probability of a transition—for
duced to elementary probabilities. As long as no consistent theory we saw that
of this couldnotlead to independent probabilities. Whatthe
probabilities was developed within quantum physics, there could theory does
be - specify, is the probability that the System-point beat a
no verdict about indeterminism. For Born, Jordan, and given place in
Pauli, this'rea- the configuration space. One might therefore hopethatthese
soning was an undercurrent of their struggles with the issueof inde- considera-
tionswould leadus to independent elementary physica
terminism. oo 4 l probabilities.”
In any case, “a trustworthy decision will only be possibl
Born, despite his openness to indeterminism, was equally open to the e after a fur-
ther analysis of quantum mechanics onthelines laid down
possibility, echoing Einstein, thata priori transition probabilities by Bornand
could Pauli” (Jordan 1927b, 569).
be viewed asa flaw in quantum theory: “It is notable that here [in quan-
Born’s probabilistic treatment had manydifficulties. I have already
tum mechanics], even historically, the concept ofa priori probability has
mentioned oneofthem: thestatistical nature of Born’s solutionin con-
played a part that could not be thrownback on equally probable cases,
trast to thestrict causality of individual atomic processes (supported by
for example, in the transition probabilities for emission: Of course, this
the Bothe-Geiger experiments). Several people, Schrédinger and Hei-
might be merely a weakness of the theory” (1927b, 9). Pauli similarly
senberg among them, began to realize the problem of interference
argued against the current form of quantum mechanics, claiming that of
probabilities (Schrédinger to Joos, 17 November 1926, AHQP;
probability assumptions should not. figure in the fundamental propo- Heisen-
berg to Jordan, 7 March 1927, AHOP). Moreover, Born’s
sitions of a theory (Pauli to Bohr, 17 November 1925, PC ). A searching electron, if a
. particle, was a strange creature indeed. Represented by a plane
appraisal of the quantum theoretical situation was made by Jordan: wave,it
had a precise momentum, but no definite position at all (Schrédinger
“We must reduce the quantum theoretical probabilities to independent
to Born, 2 November 1926, AHQP)! It was about these crucial
probabilities. Only then.can. we say that we really understandthe laws. problems
that quantum physicists addressed each otherin their correspondence.
Only then can we know exactly whatis causally determined, and what
Andtheir wide-ranging dialogues resulted in new theoretical
is left to chance.” And even more eloquent: “The circumstance that and in-
terpretive developments. Born’s probabilistic interpretation
quantum lawsare laws of averages, and can only be applied statistica wasa focal
lly point of these exchanges—between Born and Pauli, Schrédinger and
to specific eleméntary processes, is not a conclusive proof that theele-
Born, Pauli and Heisenberg—precisely because it was crying out for
mentary laws themselves can only beput in termsof probability (Jor-
elucidation, modification, and confirmation, not because it was an ob-
dan 1927b, 569). vious, uncontroversial advance.
Becausestatistical calculations are not in themselves indications of
The development of transformation theory by Jordan (1927f) and Di-
indeterminism,it is clear why none ofthe quantum physicists, though
rac (1927), and especially Heisenberg’s uncertainty paper, signified a
acknowledging the possibility of indeterminism, were ready to take
— breakthroughonthe issue of indeterminism. Heisenberg, impressed by
anyforceful stand on the issue. Born’s probabilistic interpretation of the
Jordan’s discussion, announced that he had solved definitively (defini-
stationary states of an atom wassignificant precisely becauseit pointed
| tiv festgestellst) the philosophical problem of indeterminism: Quantu
to the possibility of resolving this crucial issue. Born’s interpretation m
theory is essentially a probabilistic theory, not because quantum
changedthestatus of transition probabilities from a priori to calcula laws,
ble as opposed toclassical laws,arestatistical in Born’s sense (this
from the more elementary probabilities of the stationary states.'? possibil-
The ity is contradicted by the Bothe-Geigerresults), but because one needs
13. “Born ist mit diesen Aufstellungen in einem wesentli
chen Punkte iiber den ur-
spriinglichen Heisenbergschen Gedankenkreis hinausgegangen: Wahrend
bei Heisen- Heisenberg’s conception on an essential point: Whereasfor Heisenbe
bergals statistische Begriffe nur die Ubergangswahrscheinlichkeitener rg only transition
schienen, hat Born probabilities appearedasstatistical concepts, Born introduced
als primaren Begriff die Zustandwahrscheinlichkeit eingeftihrt, probabilityof a state as the
aus denensich die Uber- fundamentalconcept from which the probabilities of transitions are
gangswahrscheinlichkeiten erst sekundar ergeben (durch Bestimm deducible merely as
ung der zeitlichen An- secondary concepts [through determination ofthe time variation of
derungen der Zustandswahrscheinlichkeiten)” (By using these the probabilities of
ideas Born transcended states]) (Jordan 1927a, 646).
64 Chapter Three

probabilities to describe fully the state of the system. The future cannot CHAPTER 4
be knowninallits detail, not because quantum lawsarestatistical, but EX
because the present cannot be knowninall its detail and therefore must
be described probabilistically (Heisenberg 1927b). The Dialogical Emergence of
Heisenberg’s uncertainty paper made a forceful public statementin
favor of indeterminism, in contrast to the earlier tentative employment
of statistical considerations. In addition, Heisenberg’s paper restored
Heisenberg’s Uncertainty Paper
particle ontology within the atom, in opposition to Schrédinger’s
waves. Despite disagreements among Heisenberg, Dirac, Jordan, Bohr, Let’s get physical.
and Born on the ultimate source of acausality, acausality itself, from this
point forward, becamea pillar of the quantumorthodoxv: an umbrella Singer Olivia Newton-John, Max Born’s granddaughter
under which different persuasions were united, a sword with which
opponents were chastised as reactionaries. In historical accounts writ-
ten by the orthodox, indeterminism becamea “natural interpretation”
in Feyerabend's (1975) sense: indeterminism was present in quantum
theory from the very beginning; one simply had to “realize” fully what Introduction
one hadactually knownall along.**
Less than two years passed between Heisenberg’s reinterpretation pa-
14. See, for example, the discussion ofclassical indeterminism in Born (1955a). per (1925) and his formulationof the uncertainty principle (Heisenberg
1927b). During this time, Heisenberg often contemplated interpretive
issues related to the new quantum theory. How is one to interpret the
new formalism, when space-time imagery no longer seemsto fit? What
can the notion of “interpretation” in such circumstances mean? Should
one give up space and time notions, as the original formulation of the
matrix philosophy implied, or can one determine, in a systematic way,
the limits of application of the old, classical concepts? Whatis the con-
nection between quantum theory and classical mechanics, when Bohr’s
correspondenceprinciple is no longer necessary as anindependentpos-
tulate? Are the discontinuous and acausal features of the new theory
derivative, or fundamental? Does the apparentincompatibility between
the new formalism and experiment (precise times.of energetic transi-
tions and continuoustrajectories of an electron in a Wilson cloud cham-
ber that the new formalism is unable to describe) imply that quantum
theory is in need of further elaboration and modification? Or should
one accept the formalism as complete and final and undertake a suit-
able reinterpretation of experiments and experimentalterms?
Heisenberg’s elaboration of these issues took place in a contextof in-
tense dialogue.In this chapter I describe how Heisenberg’s opinion on
these central issues emerged gradually in his dialogues with Schro-
dinger, Pauli, Dirac, Jordan, Campbell, and Sentfleben.I analyze Heisen-
berg’s dialogue with Bohrin chapter 6. The emergenceof Heisenberg’s

1. Heisenberg (1927b) is quoted in translation from Wheeler and Zurek (1983). Page -
references are to Wheeler and Zurek. I have retranslated the opening sentenceofthe pa-
per (see pages 69 and 109 below) to renderit more faithful to the original Germantext.
 66 Chapter Four
The Dialogical Emergenceof Heisenberg’s Uncertainty Paper
67
interpretation, with the uncertainty principle as its main focus,
was in- before Heisenberg, suggested that the path of an intraa
timately connected with his growing confidencein the finalit tomic electron
y of the should be regarded as an irregular sequence of discrete points. Jamme
new quantum theory. This confidence crystallized as he commun r
icated has identified the two claims as the most startling and novel
with Dirac, Pauli, and Jordan and was intensified
by his emotional re- in Heisen-
action to Schrédinger’s competitive interpretive attempt berg’s uncertainty paper(1974, 85).
s. It would un- Meisenberg’ least expected interlocutor was the Germa
derpin his resolution of other issues, notably those of acausal n philosopher
ity and Johann Fichte. One of the most pregnantideas in the uncert
discontinuity. ainty paper
Heisenberg’s treatment of the space-time problem inthe quant - was Heisenberg’s reduction of wave packets during measu
um do- rement. Such
measurement selects a definite value for a measu
main similarly took place in the midst of many overlapping dialogu red observable “from
es. the totality of possibilities and limits the options for all
Schrédinger’s version of quantum theory provided both stron subsequent
g in- measurements” (Heisenberg 1927b, 74). With this
centive and a rich conceptual resource for the restoration idea, Heisenberg in-
of space-time . augurated the notorious measurement problem of quant
imagery in the quantum domain. Pauli urged Heisenberg to find um mechanics,
a which plagues physicists and philosophers of quant
particle-kinematic interpretation, drawing Heisenberg’s attenti um physics to this
on to the day. The source ofthis idea, as Heisenberg pointed out just
problem ofthejoint inapplicability of conjugate variables. Dirac a few years
singled later,.in 1932, during his lecture at the Academyof Science in Saxony
out the basic commutation relation pq — qp = (h/27i)1 as , .
the focus of was Fichte’s philosophy ofself-limitation of the ego: “The observ
ation °
a space-time analysis, despite the earlier stand taken by matrix
physi- of nature byman showshere a close analogy to the individ
cists that this formula is unamenable to visualizable interpretation ualactof
. Jor- perception which one can,like Fichte, accept as a processof the Selbst-
dan, Campbell, and Sentfleben, each in his own way, analyz
ed the con- Beschriinkung des Ich (self-limitation of the ego).” Heisenberg explic
nection between the issues of space-time and causality in ated
quantum Fichte’s idea inthe following way: “It meansthat in everyact of percep
theory. Their treatment triggered Heisenberg’s creative dialogi
cal re-
-
tion weselect one of the infinite number of possibilities and thus we
sponse, culminating in the formulation of the uncertainty principle.
also limit the numberof possibilities for the future” (1952, 28).
Heisenberg’ decision to focus his analysis on the operational meaning: These
wordsare almost identical with the concludinglines of the uncertainty
of experiments emerged only in thelast stages of his interpretive efforts,
paper: “Everything observed is a selection from a plenitude of possi--
through dialogues with Jordan, Campbell, and Dirac. In this
context . bilities and a limitation on what is possible in the future” (Heisenberg
Einstein’s dictum—“It is the theory which decides what
serve” —becamecrucial.
we can ob- 1927b, 74).
Accordingto the usual accounts, Heisenberg’s challenge in 1927
was Dialogue with Schrédinger
to demonstrate compatibility between thelimited degree of precisi
on
with which conjugate variables could be simultaneously determi
ned . Heisenberg’s need to counter the threat posed by Erwin Schrédingerto
and the limited accuracy with ‘which such variables could
be mea- the Géttingen-Copenhagen achievement, and thetheoretical possibili-
sured in experimental arrangements (Jammer 1974, 61). Yet the idea
of ties opened bySchrédinger’s work, form the background neededto un-
such a one-diznensional comparison of theory and experiment, no mat-
ter how adequate to describe the underlying logic of an argume derstand the emergence of Heisenberg’s uncertainty paper. In reaction
nt, can to Schrédinger’s challenge, Heisenberg resurrected Anschaulichkeit and
hardly do justice to the contrived process of Heisenberg’s
creative ef- the regular space-time container, eliminated by the original philosop
forts. Nor doesit suffice to elucidate the text of the uncertainty paper hy
of matrix mechanics. Heisenberg thus restored the particle-kinematic
(see chapter 5).
interpretation of quantum theory. In an emotional confrontation
The analysis in this chapter demonstrates the indispensability with
of Schrédinger, Heisenberg moved gradually, if not to a commitment,at
“lesser” scientists in the dialogical process. Twocrucial compon
entsof least to a forceful emphasis on the discontinuity and acausality of quan-
Heisenberg’s uncertainty paper were provided by
“marginal” physi- tum description. It was the threat implicit in Schrédinger’s work that
cists—H. A. Sentfleben and Norman Campbell. Sentfleben provid
ed pushed Heisenberg toward the assertion that the new quantum me-
the formulation for Heisenberg’s “definitive” resolution
of the causality chanical formalism wasfinal. While publicly denying anyphysicalsig-
issue: causality does not hold because the present cannot
be known pre- nificance to Schrédinger’s wave mechanics, Heisenberg,in
cisely and therefore the future cannot be predicted exactly. his efforts
Campbell, to create an alternative to Schrédinger, relied substant
ially on the
 68 Chapter Four
The Dialogical Emergence of Heisenberg’s Uncertainty Paper 69
“enemy's” theoretical resources and physical imagery. We often hear dinger’s theory butalso workby other physicists that substantiated the
Heisenberg speakin conflicting voices on the sameissue: while resur- wave theoretical point of view. Thus Heisenberg found Darwin's and
recting Anschaulichkeit, Heisenberg admired Unanschaulichkeit; while Wentzel’s work abscheulich (Heisenberg to Pauli, 9 March 1927, PC ). An
asserting the physical equivalence of the different versions of quantum
irritated Heisenberg complained to Pauli that physicists should also
theory, Heisenberg attempted to argue that the matrix version was nev-
learn matrix methods. Significantly, Heisenberg informed Pauli about
ertheless more significant ontologically. his interpretative efforts immediately after expressing his discontent
Restoration of Anschaulichkeit and of Space-Time with the wave theory's success (Heisenberg to Pauli, 4 November 1926
and 5 February 1927, PC). This success, Heisenberg realized, posed
The uncertainty paper represents a striking epistemological about-face a direct threat to his achievement, and Heisenberg decided to.counter
to the original philosophical interpretation of matrix mechanics. If in
Schrédinger on his own terms. , oe
the reinterpretation paper (1925) Heisenberg claimed that observable Even thoughthe uncertainty paper is knownforits refutationof clas-
magnitudes dictate the structure of a theory, his uncertainty paper
sical determinism, Heisenberg’s original concern was with the “intui-
stated that the theory determines what can be observed. In 1925 Heisen-
tive content of quantum theoretical kinematics and mechanics.” This
berg eliminatedclassical space-time and Anschaulichkeit; in 1927 he re- is, in fact, the title of the uncertainty paper. Heisenberg began this
stored them.’ In 1925 Heisenberg considered the position ofthe elec- paper with a definition of the Anschaulichkeit (intuitiveness) of physi-
tron unobservable; in 1927 he described a thought experimentby which
cal theory: “Webelieve that we intuitively understand physical theory
electron position within an atom could be measured. Heisenberg’s re-
when wecan think qualitatively about individual experimental conse-
versal wasa direct response to Schrédinger’s wave mechanics. quences and at the same time we know that. that application of the
One of Schrédinger’s arguments for the superiority of his approach
theory never containsinternal contradictions” (Heisenberg 1927b, 62).
over the matrix approach wasthe familiarity and intuitiveness ofwave By analyzing the measurementprocess with the help of a y-ray micro-
theoretical notions. Heisenberginitially dismissed such declarations as
scope, Heisenberg claimed to have demonstrated that it supported the
“rubbish” (Heisenberg to Pauli, 8 June 1926, PC). Yet toward the winter
basic commutation relation pq — qp = ~ih (1927b, 64). According to
of 1926, Schrédinger’s success was becoming manifest—a deluge ofpa-
his own definition, Heisenberg succeeded in providing a visualizable
persfollowing Schrédinger’s approach and ignoring the matrix method
interpretation of the matrix formalism: “All the concepts which can be
swelled the scientific literature. The familiar tools of wave mechanics
usedin theclassical theory for the description of a mechanical system
were being successfully applied to a great variety of problems unamen-
can also be defined exactly for atomic processes in analogy to theclas-
able to matrix treatment. sical concepts” (1927b, 68). In this way definability is not restricted by
Jordan (1927a) explained the overwhelming success of Schrédinger’s
indeterminacy relations, because each physical variable, if measured
theoryby citing not only its mathematical convenience butalso its wide
alone, can be measured with arbitrary accuracy: it is the simultaneous
intuitive appeal. Schrédinger (1952a) indeed hopedto provide an in-
measurementof physical variables that sets limits on the precision of
tuitive interpretation that would relegate Bohr’s quantum jumpsto the measurement. ~
status of Ptolemaic epicycles. No wonder neither Bohr himself northe
Heisenberg discussed other experiments and concluded: “As we
authors of matrix mechanics (erected on the basis of Bohr’s concepts)
can think through qualitatively the experimental consequencesof the
sympathized with Schrédinger’s aspirations.
theory in all simple cases; we will no longer have to look at quantum
The success and popularity of Schrodinger’s theory very much an-
mechanics as unphysical and abstract” (1927b, 82). In order to dispel
noyed the ambitious Heisenberg. He called disgusting not only Schré-
any doubt as to whose accusations he was refuting, Heisenberg ap-
pended a footnote to this statement, reminding the reader of Schré-
2. The matrices p and initially had no kinematic meaning: they were,
as Heisenberg dinger’s remark aboutthe “repulsive” and “frightening” nonintuitive-
admitted in a letter to Pauli, simply formal symbolic artifacts. Precisely
because the for- ness of the matrix theory.It is not the Unanschaulichkeit of Heisenberg’s
mulas had only symbolic meaning,originally no attempts were madeto
draw any kine-
matic conclusions from them. For example, the basic commutat
ionrelation PY - qp= theory but the misleading Anschaulichkeit of Schrédinger’s that hinders
(h/2ai)1, if interpreted kinematically means that p and q cannot simultan the progress of physics, Heisenberg said in effect.
eously have .
definite values. This meaning was contained in the theory from the very beginning Despite these statements, Heisenberg still considered lack of visual-
,yetit
wasonlya yearlater that physicists drewthis conclusion.
izability a positive feature. In a letter to Pauli, Heisenberg explained
 70 Chapter Four
The Dialogical Emergence of Heisenberg’s Uncertainty Paper
71
that he preferred Dirac and Jordan’s theory to wave mechanics because
credit the wave theoretical approach. As Heisenberg wrote to Pauli,
it was “less visualizable [unanschaulicher] and more general” (Heisen-
considerations of the transition from micro- to macromechanics were
berg to Pauli, 31 May 1927, PC). Heisenberg’s effort to render matrix
the most important ofall: first he would like to demonstrate how this
mechanicsvisualizable was undertakenless to satisfy his ownintel
lec- transition does not occur (Heisenberg to Pauli, 23 February
tual preferences and more to contribute to the matrix theory’s 1927, PC). It
wider does not occur as Schrédinger originally intended, through wave pack-
acceptance.?
ets that keep their identity.* The spreading of the wave packet, Heisen-
Against Waves - berg pointed out, constitutes a conclusive refutation of the possibility
of a micro-macro transition along the lines outlined by Schréding
In order to comprehend Heisenberg’s reasoning in: the uncertainty pa. er.
Norcanthisdifficulty be avoided through considerationsof natural ra-
per, we have to take into account his hostility to Schrédinger’s wave
diation width (Strahlungsbreite): the transition from quantumto classi-
theory as a competing system, and his resulting aversion to wave
theo- | cal mechanics mustbe intelligible without the use of electrodynamics
retical concepts. Heisenberg himself regarded as “the most important
(Heisenberg 1927b).
result of the uncertainty paper” his critique of Schrédinger’s idea of
the Heisenberg’s change of heart was complete. If in 1925 he attempted
wave packet (Heisenberg to Pauli, 23 February 1927, PC). to reinterpret the mechanical behavior of an electron by replacing kine-
One of the problems Heisenberg treated in his uncertainty paper matic concepts with electrodynamic ones (elementary waves of emitted
wasthe transition from micro to macromechanics. A description ofthis radiation), now he refused to employ electrodynamicsin the physical
transition was essential for obtaining a physical interpretation of the
interpretation of quantum mechanics. Upon coming to work with Hei-
quantum formalism. Yet using the matrix theory, Heisenbergfailed re- senberg, Felix Bloch had proposed to explore the possibility of keep-
peatedly to provide suchadescription, and Schrédinger seized onthis ing wave packets together through radiation damping. Heisenberg, as
failure to proclaim the superiority of his theory over the matrix version. Bloch recalled, was strongly opposed to this initiative. “As to my
Schrédinger attempted to identify microphenomena with wavevibra- hopesfor keeping wavepackets together by radiation damping, he only
tions and succeeded in constructing the example of the harmonic oscil- smiled and said that, if anything, it could of course make them spread
lator, in which superposed vibrations constitute a well-defined wave
even more” (Bloch 1976, 25; interview with Bloch, AHOP).
packet(a configuration of waves that does not spread in time, and that There are other reasons for regarding Heisenberg’s paperas a direct
therefore keepsits identity; Schrédinger 1926e). Transition to the macro- attack against Schrédinger’s competitive efforts. Only under such an
level is achieved bynoticing that this wave packet can be identified assumption do some puzzling features of Heisenberg’s paper, as well
with an oscillating particle. Schrddinger also hoped to construct wave as the story of Heisenberg’s clash with Bohr, become intelligible (see
packets that would move along Keplerian ellipses and therefore repre- chapter 6). In his description of the y-ray thought experiment, Heisen-
sent electrons movingin the atom. Schrédingerrealized early, however, berg had committed a trivial error, which both Bohr and Dirac (who
:
that such a direct intuitive interpretation of the y-function would not wasin Copenhagenatthe time) broughtto his attention (Heisenberg to
be feasible. He subsequently interpreted y(r) as a weight. function and Pauli, 16 May 1927, PC). Heisenberg had treated both photonsandelec-
stated clearly that the y-function “cannot and may notbe interpreted trons as regular point particles and arguedthatat the timeof their col-
directly in terms of 3-dimensional space—however much the one- lision a photon transfers to an electron a discrete and uncontrollable
electron problem tends to mislead us on this point” (1926d, 120).
amount of momentum (Compton recoil). The more precisely the posi-
Thus Schrodinger abandonedhisfirst, “naive” interpretation and de- tion of the electron is determined, the greater the uncertainty of the
clared this in print a few months before Heisenberg’s uncertainty
paper discontinuous changein the electron’s momentum (Heisenberg 1927b,
was published. That Heisenberg found it appropriate to attack a
posi- 64). Yet Compton recoil does not lead to indeterminacy but rather to
tion his opponentno longer held is indicative of his eagerness to dis-
~ exactly calculable momentum changes. From the conservation lawsfor
3. Anschaulichkeit, as Forman (1979) and Miller (1978) have
argued, was a pregnant 4. Were the wave packets to carry the periodic motionsof classical electrons, then the
notion in the German cultural milieu. In the German neo-Kantian philosoph
icaltradition, atom would send out radiation that could be expandedin Fourierseries, for whichthe fre-
it was important as a cultural value. Forman and Miller have interpret
ed Heisenberg’s quencies of the harmonicsare integral multiples of the fundamental frequency. “The
restoration of Anschaulichkeit as a submission to the pressure of a widerinte
llectual mi- frequencies[of] the spectral lines sent out by the atom are, however, according to quan-
lieu. Yet Heisenberg showedlittle desire to tackle the issue of Anschaulic
hkeit, until sub- tum mechanics, never integer multiples of the basic frequency — except in the special case
jected to concretescientific press ure—by Schrédinger.
of the harmonicoscillator” (Heisenberg 1927b, 73).
 72 Chapter Four
The Dialogical Emergenceof Heisenberg’s Uncertainty Paper 73

energy and momentum, and knowing the energy and momentum of

That a physicist of Heisenberg’s stature would make such a mistake
the colliding photon, one can calculate the momentum change of the
electron exactly. There is no way to transcendtheclassical deterministic
is odd enough,buthisrefusal to correct the mistake, despite powerful
criticism from Bohr, becomes incomprehensible unless he had a vested
framework onceit is assumed that photons and electrons are point par-
interest in preferring a misleading description to the correct one. Schré-
ticles obeying conservation laws (see diagram).
dinger sought to argue that reality consists of waves, and wavesonly;
Heisenberg was nowtaking the opposite stand—thatreality consists of
particles, and particles only. This suddeninsistence on the exclusively
corpuscular nature of matteris all the more surprising if we recall that ©
only three months earlier Heisenberg (1926b) asserted that recent de-
velopments in physics (matrix theory and theBose-Einsteinstatistics)
undermined thereality of particles. It was the particle characteristicsof
radiation and matter that Heisenberg now wanted to emphasize, at the
expenseof their wave characteristics. Heisenberg’s letter to Pauli clearly
identifies the focus of his disagreement with Bohr: “I argue with Bohr
about howfarthe relation ApAq ~ h hasits origin in the waveor discon-
tinuity aspect of quantum mechanics. Bohr emphasizesthatthe diffrac-
tion of wavesis essential in the y-ray microscope; I emphasizethat the
light quantum theory and even [the] Bothe-Geiger experiment[s] are
essential” (Heisenberg to Pauli, 4 April 1927, PC).
There is an important phrase in-the passage above: “discontinuity
aspect of quantum mechanics.” Heisenberg came to regard disconti-
nuity as the mostessential feature of the quantum world. The matrix
theory, and not wave mechanics, was best suited to express this basic
discontinuous aspect (Heisenberg.to Pauli, 16 May 1927, PC). After
completing the uncertainty paper, Heisenberg wrote to Jordan that “fi-
nally” he felt the discontinuous ground firmly under his feet (Heisen-
berg to Jordan, 29 September 1926, AHOQP). Clearly, replacing his own
description based on the discontinuous Compton recoil with Bohr’s ex-
planation based on the continuous waveaspect of radiation did not suit
Heisenberg’s purposes. As Heisenberg wrote to Pauli, discontinuity, as
the hallmark of the quantum world, could not be overemphasized; he
wastherefore happy with his paper, despite the mistakes (Heisenberg
to Pauli, 16 May 1927, PC). . :
Bohr, as one gathers from Heisenberg’s postscript to the uncertainty
paper, also pointed out to Heisenberg that Compton recoil appliesrig-
orously only to free and not to boundelectrons. And for free particles
Bohr’s ownderivation of the uncertainty relations, from the wave theo-
In his initial mistaken analysis of a y-ray microscope, Heisenberg treated both photons
and electrons retical point of view, can be considered more natural. Yet Heisenberg
as particles without taking into accounttheir waveattributes. This analysis is
equivalentto the scat- preferred the Dirac-Jordan quantum theory, which was moresuited to
teringof light quanta byintraatomicelectrons in the Compton effect (bottom),
andit leads notto inde- express discontinuities. This was the reason he ended up “in the camp
terminacybut to exactly calculable changes. Heisenberg’s misleading analysis
is sometimes of matrices and against waves” (Heisenberg to Pauli, 31 May 1927, PC).
repeated in popular expositions of uncertainty principle.
Oskar Klein, a witness to Heisenberg’s clash with Bohr, offered the
 74 Chapter Four The Dialogical Emergence of Heisenberg’s Uncertainty Paper 75

following commentary: “It was so muchin the beginning,... although ematical version of quantum mechanics, the Heisenberg uncertainty
Heisenberg must have knownin principle, if one had asked him,that
principle, and Bohr complementarity.
there couldn’t be much difference [between particle and wave theoreti-
Yet on positivist grounds, the issue was meaningless. If the two theo-
cal descriptions]. . . . Still, I think, he was not so familiar with the wave
ries were fully equivalent, as the participants seemed to hold, what was
point of view at the time, and he wasa bit partial to the matrix point
the meaning of the claim that one theory was superior to the other on
of view which hestarted” (interview with Klein, AHOQP). Heisenberg,
physical grounds? Howcould anyone defend the position, as Heisen-
as he revealed in an interview with Kuhn (AHQP), was inspired by
_berg did, that his theory was able, though Schrédinger’s equivalent
William Duane’s (1923) results, which managed to explain the diffrac-
theory was not, to deduce certain physical results, such as Planck’s
tion of light by a grating assuming the corpuscularnatureof light (light
law? And what meaning could one attach to a “real existence,” and in-
quanta) and avoiding wave theoretical presuppositions(see chapter11). stantaneity, of the unobservable quantum jumps,the reality of which
Heisenberginsisted in the uncertainty paperthat the powerfultransfor- _
Heisenberg was eager to demonstrate? The whole controversy gains in-
mation theory allowed a purely corpuscularinterpretation in the spirit ~~
telligibility only when we assume that not only Schrédinger but also
of the generalization of Born’s statistical interpretation. He therefore re- Heisenberg had someverystrong reason to ask how unobservable pro-
jected not only Schrédinger’s attempts at a pure wave ontologybut also
cesses really occur in nature.
Bohr’s conciliatory solution in termsof wave-particle duality (see chap-
Before the wave theory appeared in print, Heisenberg spoke with
ters 6 and 11). eT . A "two different voices on the issue of continuity versus discontinuity. Hei-
" Nevertheless, Heisenberg supported Bohr’s position on complemen-
senberg incorporated the guidelines of Born’s approach to building “a
tarity in the postscript to the uncertainty paper. In orderto offer a uni-
truly discontinuous theory”in his reinterpretation paper, by integrat-
fied explanation, capable of countering the opposition, differences of
ing the concepts of discrete energy levels and transitions between them
opinion were to be suppressed. The easiest way to deduce the uncer-
into the basis of the matrix mathematical formalism. At the time Hei-
tainty relations is, in fact, by identifying a particle with a limited wave
senberg did not claim (as he would later, after the appearance of con-
field, as Bohr argued in his Comolecture. Heisenberg; who consid-
tinuous wave mechanics) an irreducible, a priori status for disconti-
ered the uncertainty formula fundamental to the self-consistencyof the
nuities. In fact, Heisenberg was elated (as Schrédinger would later
quantum formalism, was willing to employ Bohr’s accessible derivation
be) when helearned that his new mechanics did not have to postulate
for wider audiences and for “pedagogical reasons.” He used it because
discontinuous energylevels—their existence simply followed from the
“it did not do any harm” to his own explanation, yet he did notbelieve
mathematical formalism. As Heisenberg noted, nothing in the matrix
“it was necessary”(interview with Heisenberg, AHQP).
theory depends on assuminga priori that discrete stationary states ex-
Discontinuities and Quantum Jumps ist—boththe discrete and the continuous solutions follow equally from
the basic quantum equation: pq — qp = (h/27i)1 (Heisenbergto Pauli,
In the heated debate between Schrédinger and the matrix physicists,
18 September 1925, PC).
each side proclaimed its own approach superior for. describing atomic
Similarly, Heisenberg, Born, and Jordan argued (as Schrédingerlater
phenomena. Schrédinger claimed that wave mechanics was better be-
would abouthis theory) that “a notable advantage” of the new matrix
cause it operated with familiar and manageable mathematical notions
theory “consists in thefact that the existence of discrete stationarystates
and promised toeliminate the irrationalities of quantum jumping. The
is just as natural a feature of the new theory as, say, the existence of
matrix physicists considered these “advantages” superfluous and mis-
discrete vibrations frequencies in classical theory” (Born, Heisenberg,
leading—it was the matrix theory and not Schrédinger’s, they argued,
Jordan 1926, 322). Although Bohr’s concepts played a heuristic role on
that expressed the “inevitable,” irreducible discontinuities of the quan-
the wayto the developmentof the matrix theory, the matrix theoryitself
tum world.
EE undermined the fundamentalrole of a priori stationary states and “ir-
The confrontation between Schrédinger and his opponents is one of
reducible” quantum jumps (chapter 2).
‘ the mostfascinating chapters in the history of physics. Although emo-
= These concepts wereresuscitated in responseto the threat from Schré-
tions ran high (or maybe because they did), the dialogue that resulted
dinger. Matrix physicists, unable to develop a comprehensive andsatis-
producedsignificant contributions by all participants.It led to the Born-
Pauli probabilistic interpretation, the Dirac-Jordan generalized
fying interpretation of quantum physics, and faced with Schrédinger’s
math- rival theory, chose the only effective strategy available— claiming that
 76 Chapter Four
The Dialogical Emergence of Heisenberg’s Uncertainty Paper 77

an interpretation dissociated from the original G6éttingen-Copenhagen In Heisenberg’s fluctuation paper (1927a), the formulas are correct,
approach was impossible, that quantum discontinuities could not be
but Heisenberg’s logic is faulty. If we designate by D, discontinuous
eliminated. jumps, M, the matrix version of quantum mechanics, and F, the fluc-
Heisenberg and Schrédingerfirst clashed over the interpretation of tuation formulas, then Heisenberg’s incorrect reasoning runs as follows:
quantum mechanics at the Munich conference in July 1926, where D5 Fand M -> F;therefore M > D.
Schrodinger waspresenting his new theory. Most of the participants
Because in using matrix mechanics Heisenberg restricted himself to
were captivated by Schrédinger’s mathematics and in sympathy with only the noncontroversial assumptions of that approach—namely,that
his interpretative aspirations—at least according to Heisenberg’s as- a diagonal elementof a matrix represents the time average of the.corre-
sessment. The next confrontation-occurred in Copenhagenin the fall of sponding physicalvariablein the stationary stateconsidered—hefelt he
1926, when Bohrjoined forces with Heisenberg to combat Schrédinger’s
had proved conclusively the inevitability of quantum jumps. In orderto ©
attempt to do away with quantum jumps.
do so, however, Heisenberg had to present amuchstronger case. Heisen-
In order to counter Schrédinger, Heisenberg had to argue that dis-
berg had to provethat everypossible continuous description would lead
continuities are irrevocable.A natural candidate for this task offered inevitably to predictions incompatible with experimental results or to in-
itself— fluctuation phenomena.Einstein had analyzedfluctuation phe-
ner contradictions. Heisenberg did not even attempt this impossible
nomena to deducethe existence of discontinuities on the microlevel.
task. (And, in fact, Bohm has recently offered a continuous description
He ascribed Brownianfluctuations to the atomistic structure of matter,
of quantum jumps; Bohm and Hiley 1993). What Heisenberg did show,
while the fluctuations of energy and momentum of gas molecules led not surprisingly,is that the assumption of quantum jumpsis admirably
to his idea of light quanta. Heisenberg chose to treat fluctuations pre-
compatible with the matrix formalism. Heisenberg arguedin letter to
cisely because he viewed them as an appropriate tool with which to Jordanthat it was the matrix formalism, and not Schrédinger’s wave me-
emphasize discontinuities. chanics, that wasbest able to treat quantum discontinuous phenomena
In his “fluctuation” paper (1927a), Heisenberg considered two simi- ° (Heisenberg to Jordan, 29 October 1926, AHQP).
lar atoms having energies E,, and E,,. When these atoms are coupled
This letter to Jordan demonstrates how Heisenberg’s preference for
there is “a resonance”—each atom periodically changesits state from discontinuity emergedfrom his ‘opposition to. Schrédinger. Accord-
E,, to E,, and vice versa. The exchange of energy between these atoms ing to Heisenberg, the progress achievedin his fluctuation paper had
could be viewedeither (1) as due to energy jumps E, — E,, via photons nothing to do with his calculation of energyfluctuations, which was a
or (2) as a classical interaction between twooscillators. By translating simple problem that could easily have been solved by thermodynamic
the second alternative into matrix language and pointing out that for considerations on the assumption of quantum jumps (Heisenberg to.
such a system only the time averageof each individual atom is observ- Jordan, 29 October 1926, AHQP). The progress rather consisted in the
able, Heisenberg obtained the value WE, + E,,) for each atom, describ-
ability of matrix mechanics to calculate these fluctuations without the
ing thefluctuationof an arbitrary function of energy asf(E(t)) = “Cf(E,,)
explicit assumption of quantum jumps.> The important conclusion
+ f(E,,)). The only waythis formula can be correctforallfand E, argued was not that quantum jumpsareirreducible but rather that the matrix
Heisenberg, is if the energy changes discontinuously, spending equal theory was the formalism most appropriate to representing disconti-
amounts of timeat the values E,, and E,,.
nuities and quantum jumps. The matrix formalism and the existence of
Schrédinger did not wait long to meet Heisenberg’s challenge. Soon discontinuities, reinforcing each other, undermined, according to Hei-
he published a response, in which he treated the same problem on his senberg, Schrédinger’s claims for the wavetheory.
ownterms, without assuming discrete energy levels, and using instead
Heisenberg’s contention that he could prove the inevitability of dis-
only the language of frequencies (Schrédinger 1927p). Accordin
g to continuities in the domain where the matrix and wave mechanical meth-
Schrédinger, the interaction is controlled by a law of resonanc
e that re- ods were equivalent was unreasonable. Onthis issue, Schrédinger’s
quires the difference between two eigenfrequencies in one
system to be
equalto the difference in the other system. Consequently,
he succeeded
in showingthat theinteraction can be adequately described 5. “The progressconsists rather in the fact that quantum mechanics allow’s the calcu-
as a con-
tinuous change in the amplitudes ofthe four corresponding charact lation of these fluctuations without an explicit assumption of quantum jumps, but on
er- the basis of relations between q, q’ and so on” (Heisenberg to Jordan, 29 October 1926,
istic vibrations.
AHOQP).
 78 Chapter Four
The Dialogical Emergenceof Heisenberg’s Uncertainty Paper
79
position was moresensible: Schrédinger assumed the superio
rity of his Dialogue with Pauli
method because of what he saw asits greater suitability to
extension
beyond the domain where the two methods proved equivalent
(Schré- Wolfgang Pauli was Heisenberg’s closest scientific collab
dinger 1926f). orator and
critic. Heisenberg’s major breakthroughs—both the reinter
Heisenberg’s attempt to prove the inevitability of discont pretation
inuities in and “the uncertainty papers—would be unimaginable
his fluctuation paper was intended less as a genuine without Pauli’s
contribution to prior penetrating analysis of the issues fundamental to quant
science and more as a weapon to be used against Schrédinger’s um theory
theory. at the time. It was not merely that Pauli was generous withhi
In a letter to Pauli, Heisenberg apologized that he had submit s insights
ted for in his correspondence with Heisenberg; the very formul
publication a paper that contained nothing new. He explain ation of the is- |
ed that sues occurred when Heisenberg and Pauli addressed
the purpose of the paperwas purely “pedagogical”—it was directe each other. Pauli
d and Heisenberg complemented each other. It was
against those who favored the continuous approach(Heisenbe Pauli, unusually
rg to gifted at elucidating fundamentalissues, who often provid
Pauli, 4 November1926, PG)s . .
ed the essen-
eg tl tial pieces of the puzzle. It was thebrilliant “unphilosophical”
Shortly after Heisenberg’s fluctuation paper, Jordan publis Heisen-
hed a pa- berg (Pauli to Bohr, 11 February 1924, PC) who, cheerfully sustai
per alongsimilar lines (Jordan 1927e). Jordan, more consistently, did ning
conflicting inner voices, provided the epoch-making solutions to
not claim that he had proved the inevitability, or even the prefere the
ntial puzzles Pauli defined. In his later years Pauli regretted. his own
status, of discontinuities. Neither did he assert the Superiority of matrix “con-
servatism,” feeling perhaps that Heisenberg’s intellectual boldness (if
mechanics over Schrédinger’s continuous theory. He merely attempted
not recklessness) reaped richer scientific harvest and widerscientif
to demonstratethat the Géttingen-Copenhagen approach, based ic
on fame (Mehra 1976).
quantum jumps, wasnotinferior to Schrédinger’s intuitive contin
uous Pauli’s impactin his dialogue with Heisenberg overthe issues
framework. Jordan arrived at a sound conclusion: “Both theorie of the
s and uncertainty paper wascrucial in several respects. It was Pauli
their conflicting interpretations, despite their great difference, give the who in-
sisted, and eventually succeeded in convincing Heisenberg,
sameresults for all empirically testable claims” (1927e, 661).7 that the
physical interpretation of quantum mechanics must be undert
. , Dirac and Jordan’s transformation theory was based on probabilistic aken in
particle-kinematic terms, as opposed to the initial philosophy
axioms that presuppose discontinuity. This probabilistic basis of matrix
served mechanics. Pauli formulated the issueof the impossibility of the joint
Heisenberg directly in his derivation of the uncertainty princip de-
le. Dis- termination of position and momentum as the focus of interpretive
continuous quantum jumps and acausality became entrenched both in ef-
forts. Pauli also provided Heisenberg with the example of
the quantum formalism andin its interpretation to such an extentthat the time-
energy uncertainty relation—an example that Heisenberg repeate
they appearedto be “inevitable,” “hard” facts of nature. d and
reanalyzed in his uncertainty paper. By analyzing Born’s collisio
n treat-
6. “The enclosed note aboutfluctuation phenomena
ment, Pauli translated Born’s results into matrix language, thus under-
I am sending to you, not without
a longlist of excuses—that I will publish such a thingatall that actually mining claims by Schrédinger and Born that wave mechanics was
contains nothing supe-
newfor you or otherenlightened physicists, that I myself do not rior to matrix mechanics. This analysis was a crucial contrib
ution to the

ii
knowwhat to do with
it, either, etc. The reason that I wroteit is really only a pedagogi
cal one against the Lords elucidation ofthe physical interconnection between the wave and matrix
of continuums theory” (Heisenberg to Pauli, 4 November 1926,
PC). : approaches, preparing the groundforthe generalized version of quan-
7. This conciliatory mood did notlast long—Jordan soon became
the most militant tum mechanics proposed by Dirac and Jordan. In direct dialogue
opponent of Schrédinger’s interpretive efforts and an ardent defender
of quantum jump- with
ing. In his fluctuation paper Heisenberg was eager to convince himself Pauli, Heisenberg’s preference for discontinuities was strengthened,
and his audience the
that he had provedthe inevitability of discontinuities and the “real
existence” of quantum product ApAgq wasdefined, and the causality issue became morefo
jumps. Years later, Heisenberg would remember his emotional need
cused.
more vividly than
Jordan’sinitial balanced appraisal: “The result decided clearly Particle-Kinematic Interpretation
in favor of the quantum
jumpsand against the continuous change” (1977, 4). Similarly
, Dirac, whoat the time was With the benefit of hindsight, one can argue that the uncertainty
still hostile to Schrédinger’s theory, also confused the desired with
the proven. In an im- prin-
portant paper, stimulated by Heisenberg’s fluctuation considera ciple was contained in the new quantum mechanicsfrom the beginn
tions, Dirac wrote: “Hei- ing,
senberg had shownthatthese calculated meansare just what one implied by the basic commutationrelation Pq — qp = (h/2z7i)1.
would expect from the The
assumption that the energy changes discontinuoasly
from one quantized value to an- reason this fact was not mentioned and elaborated from thestartis
other. ... The theorycan thus be considered to showthat the energya that
ctually does change originally the matrices p, q, and the like were considered
discontinuously” (1927, 622). formal sym-
bolic artifacts, which had no connection with the usual meanin
g of po-
80 Chapter Four
The Dialogical Emergence of Heisenberg’s Uncertainty Paper 81

sition and momentum.It seemed a marvel, as Lorentz wrote to Schroé- expressed the opinion that one must not only eliminate theoreticaliy
dinger, that the equations of motion “can be satisfied when one unobservable entities but also endow experimentally observable enti-
interprets these symbols as things that have quite another meaning, and ties with the appropriate theoretical representation. At the same time
only remotelyrecall those coordinates and momenta” (Przibram 1967). Pauli, a lifelong critic of field theories, soughtto transform the original
Letus recall that in the original matrix approach, the matrix q was en- eledtromagnetic underpinnings of the matrix theory into a particle-
dowed with electromagnetic, notkinematic, meaning. kinematic foundation. His aim was to base the interpretation of matrix
In Dirac’s version of q-numbers, constructed in close analogy with elements on kinematic attributes of particles rather than on elementary
classical mechanics, it is appropriate to ask what happens when one waves of emitted radiation. Pauli realized that he could partly achieve
substitutes a c-numberfor a q-number(q-numbers are symbols thatsat- this objective by interpreting Schrédinger’s y-function as denoting the
isfy quantum theoretical equations; c-numbers are ordinary numbers). probability of position. -. es Ree
Onecan subsequentlysee thatit is impossible to assign c-numbersto Pauli defined |w(q1,. . . , q¢)|?dq.-- -dq, as the probability that for a
both position and momentum simultaneously without violating the system in a definite stationary state, the coordinates q, of the particles
equation pq — gp = (h/2z7i)18 The fact that one cannot talk about the of this system (k = 1,..., f) lie between q, and q, + dq,.. Pauli subse-
position and the momentum of a given atomic object simultaneously, quently interpreted the diagonal elements of the matrix of an arbitrary
and that one cannot define a path of a particle, was also implicit in function F such that Fin = SF(q:)|Wa(qu)|2dqz- - dg, as the average value
Schrédinger’s version of quantum mechanics. Schrodingerstressed that . of Fin the state m (Pauli to Heisenberg, 19 October 1926, PC):
quantum mechanical phenomena cannotbe described by image points, According to Pauli, the probability |y(q1, - . ., q,)|?dq1-- -dq,; should
only by wave processes, and soughtan interpretation of wave mechan- be regarded as observable in principle. Despite the fact that, histori-
ics analogousto the interrelation between wave optics and geometrical cally, matrix elements were connected with the properties of emitted
optics. Born’s collision treatment, in agreementwith this approach,pre- waves, Pauli became convinced with “all the fervor of his heart” that
sented the moving electron by a plane wave e~*, wherethe value of the . Matrix elements must be connected with “kinematical (perhapsstatis-
momentum p = ik is determined exactly while the position of the mov- tical) data” for particles (Pauli to Heisenberg, 19 October 1926, PC).
ing corpuscle is completely undetermined. Pauli,.as I notedearlier, after initially ignoring Schrédinger’s theory,
Pauli, who opposed Schrédinger’s interpretive ambitions and sought recognized its fundamental importance. Because of his aversion to con-
to endowthe quantum formalism with particle-kinematic meaning by tinuous field theories, Pauli had no use for Schrédinger’s interpretive
generalizing Born’s collision arguments, tried to understand why in aims, but he was not prepared to overlook Schrédinger’s physical in-
cases when momentum is known, positions is completely undeter- sights. Thus Pauli’s kinematic interpretation to a considerable degree
mined. Pauli discussed this issue in a remarkable letter to Heisenberg simply translated into particle language Schrédinger’s interpretation of
(Pauli to Heisenberg, 19 October 1926, PC). In this letter Pauli elabo- # as a measure of charge density. Thefact that Schrédinger’s y-function
rated other cases of quantum peculiarities that were incomprehensible operated in a multidimensional configuration space strengthened Pauli’s
from the classical point of view.° He also explored the question of the predisposition to ascribe to it a particle-kinematic meaning (Pauli to
observability of electron motion. On more than one occasion Pauli had Heisenberg, 19 October 1926, PC). 7
At first, Heisenberg vacillated. He was by that time ready to accept
8. “The general question of classical mechanics can be formulated as follows: Whatis
that matrix elements could be defined kinematically without recourse
the value of any constantof integrationg of a given dynamical system for any giveninitial
conditions, specified by numerical values qo, po, say, for the initial coordinates and mo- to emitted electromagnetic radiation. Yet he disagreed with Pauli’s at-
menta qn, fs? - .- [As opposed to classical mechanics] on the quantum theory one can tempt to connect such a definition with a single stationary state. The
also obtain an expression for g as function of the 9u, Po but the go and p, do not satisfy kinematic variables, suggested Heisenberg, should be tied to two dif-
the commutative law of multiplication, so that if one substituted numerical values for ferent stationary states. In this way, the question of position would in-
them the result would in general depend onthe order in which they were previously
arranged. One can thus give no unambiguous answerto the question on the quantum
tertwine naturally with the possibilities of transitions (Heisenberg to
theory”(Dirac 1927, 623). Pauli, 28 October 1926, PC).
9. For example, if a mass with energy E is directed toward a finite potential barrier
10. That the phrase “perhapsstatistical” is in parenthesis is not accidental: Pauli did
with energy V(x), when E > V,,,, the classical particle will always run over the barrier
while the quantum particle will sometimes be reflected, with a calculabl not yet hold statistical concepts to be fundamentaland therefore did not intentionally
e probability look for a statistical interpretation; it was the kinematic aspect of the problem that was of
(Pauli to Heisenberg, 19 October 1926, PC).
prime importance to him.
 82 Chapter Four
The Dialogical Emergence of Heisenberg’s Uncertainty Paper
83
Pauli’s arguments for the equivalence of description in momentum
divided by the life-time” (interview with Klein, AHQP). In formula
space and thatin position space made a great impression on Heisen- s,
this means AEAt ~ h. The connection between the breadth of spectral
berg (Pauli to Heisenberg, 19 October 1926, PC).1! Waves in momentum
lines and thelifetimeof stationary states was at the heart of the Bohr-
space, Heisenberg concluded, have as great a physical significance as
Kramers-Slater theory. According to John Slater’s memoirs (1975b), he
waves in position space (Heisenberg to Pauli, 28 October 1926, PC). The
wasthe first to raise this point in connection with Bohr’s assumption
equation gp — pq = hi thus always correspondsin the wave presentation
about radiation emitted during instantaneous transitions from one sta-
to the fact thatit is impossible to speak of amonochromatic wave at a
tionary state to another. This assumption was disposed of in the Bohr-
fixed pointin time. But if one makes the line not too sharp, and the time
Kramers-Slater theory:here, as in classical physics, radiation (albeit
interval not too short, then theidea begins tomake sense. Analogously, vir-
tual) was emitted continuously during the entire lifetime of an excited
one cannottalk aboutthe position of a particle moving with a definite
state, and not during transitions. This theoretical change took place,
velocity. Yet when one does not applythese concepts too rigorously,it
according to Slater, because he hadinsisted that the assumption of ra-
may well make sense. One catches a glimpse therefore of whyin the
diation width during extremely short transitions is inconsistent with
macrodomain these concepts can be used as approximations (Heisen-
experiment.
berg to Pauli, 28 October 1926, PC). At this stage, however, Heisenberg
Heisenberg was familiar with the Bohr-Kramers-Slater theory, and
did not seekto interpret this insight in terms ofthe limit of precision of with its underlying rationale. There was nothing exceptionalin the fact
kinematic concepts as applied to an individual particie. suggesting in- that the frequency spectrum ofa finite wave train dependsonthe time
steada statistical meaning for space-time.! during whichthe train is emitted. Only when, after the uncertainty pa-
a

Heisenberg’s discussion was,in fact, a corpuscular improvisation on per, At signified, not simplya “finite precise intervalof time,” but rather
Pe the idea of the wave packet—an idea advanced initially by de Broglie “the breadth of imprecisionin evaluation of a point of time,” did AEAt
ES and Schrédinger, and used extensively by Bohr in his Comolecture (see ~ h, like ApAg ~ i, becomea relation that signified a break with the
iS
chapter 6). The Géttingen-Copenhagen physicists obtained important past. As Slater mentioned: “It was only someyears later, in 1927, that
:
E insights by turning the rival wave imagery to-their own ends. At this
Heisenberg publishedhis paper on [the] uncertainty principle,and that
_-point in Heisenberg’s letter, Pauli commented in the margin: “Ina time- he and Bohr madethis such a great feature of their version of wave
: interval, short as compared with the period (of revolution), it is also mechanics” (1975b, 19). Not surprisingly, Heisenberg wasafraid that
meaningless to speak of a precise energy value” (PC, 352). It is this
Jaypee rier fame

his considerations in the uncertainty paper might be regarded as “an

point that led Heisenbergto the formulation of the time-energy un- old snow”(alter Schnee; Heisenberg to Pauli, 23 February 1927, PC).
certainty relation. In this endeavor, Heisenberg relied substantially on
It is clear that the relation AEAt ~ % was obtained before ApAqg ~ h.
Pauli’s (1926a) “encyclopedia”article, where the whole issue was com- Heisenberg translated AEAt ~ h into ApAg ~ f. He daringly intro-
prehensively analyzed. , duced wavetheoretical peculiarities exactly where they seemed outof
Time-Energy Relation place: the particle-kinematic framework. Pauli’s influence wascrucial.
It was Pauli whoinsisted onthe insufficiency of field concepts, and who
According to Klein, Bohr posited “the first instance of the indetermi- sought to reinterpret the matrix formalism in particle-kinematic terms.
nacy principle—thata stationary state cannot be better defined than h In this connection Pauli struggled with thestatistical meaning of the
11. These considerations also made a great impression on Born, chilling his initial
“path” of a particle, preparing the ground for Heisenberg’s discussion
preference for Schrédinger's visualizable interpretation.In a letter to Schrédinger (6 No-
13. As Slater recalls: “Thedifficulty [with Bohr’s theory] was with the assumptionthat
vember 1926, AHQP), Born informed him of Pauli’s calculations in the case of a two-
the radiation was emitted in the form of a photon at the instant the atom jumped from
dimensional rotator. Its solution can be written in two ways:first, viewing x, y as regular
the state E, to the state E,. Any student of physics or mathematics knows that a wave
coordinates and p,, p, as operators; second, viewing p,, py as variables and x, y
as opera- train of finite length has a frequency spectrum which is not strictly monochromatic, but
tors. Both solutions lead to the sameresult. The possibility of substituting momentum
which instead has a frequency breadth Av whichis of the order of magnitude of 1/T,
space for coordinate space (through an appropriate canonical transformation) reveals—
where is the length of time during which thetrain is emitted... . If this period of time
Pauli, Born, and Heisenberg concluded—that there was no reason to supposethat q-space
is much longer than the period ofoscillation, which is 1/y, the breadth Av will be very
has any conceptual preference andthat the essence of quantum phenomena dependson
small compared to the frequency v. The observed sharpness of the spectral lines
the visualization in q-space (as Schrédinget believed). shows
that this must be the case. The experiments are consistent with emitted wave trains which
12. On this point, see the discussion of Heisenberg’s dialogue with Campbell, later
in have perhaps the order of 105 or more waves on the train. ... Surely it must have
this chapter. taken
long enough for 105 wavesto be emitted” (1975b, 10).
 The Dialogical Emergence of Heisenberg’s Uncertainty Paper 85
84 Chapter Four

of this problem in the uncertainty paper. Pauli explained his approach tum jumps’ must be as concrete and determinable by measurementas,
in this way: if at time f the particle's position is qo, what is the proba- say, energies in stationary states.” The imprecision within which the
bility that at time f + 7 its coordinate will be q? The kinematic answers, instant of transition is specifiable, argued Heisenberg,is given byAf in
argued Pauli, should be statistical, so one would not be able to talk AEAt ~ h, where AE indicates the change of energy in a quantum jump
about the definite “path” ofa particle, quite in accordance with the im- (1927b, 76). Analyzing the transition from a stationary state n = 2 into
possibility of endowinga particle simultaneously with exact position n = 1and using a thought experimentsimilar to the Stern-Gerlach ex-
and momentum (Pauli to Heisenberg, 19 October 1926, PC). eriment for energy measurement, Heisenberg argued that in order
In his uncertainty paper Heisenberg discussed several cases of the to be able to make a distinction between the twostates, At cannot be
time-energy uncertainty relation, includingthe case of the Stern-Gerlach smaller than h/ AE. The instantof transition, Heisenberg concluded,.can
experiment and that of transitions from onestationary state to an- only be determined within a At-spread (1927b, 77). Heisenberg inter--
other.* These examples were treated extensively in Pauli’s (1926a) preted Af as characterizing the imprecision of thedefinition of time,
important encyclopediaarticle, and Heisenberg dutifully appended a rather than as a finite, precise interval of time. Discussing the Stern-.
footnote referring the reader to Pauli’s work: “Quantum mechanics has Gerlach experiment, Heisenberg demonstrated that the “accuracy of
changed only slightly the formulation of these problems as given by the energy measurement decreases as we shorten the time during
Pauli” (1927b, 63 n). ; which the atom is underthe influence of the deflecting field’ (1927b,
In his encyclopedia article Pauli discussed the problem of energetic 67).5 Heisenberg interpreted ¢ ; (during which the atoms are under the
transition from one stationary state to another, referring to Einstein’s
influence of the deflecting field) as “uncertainty in time,” rather than
treatment of elementary processes of absorption and emission in proba- as a short, precisely defined interval of time. In this way, Heisenberg
bilistic terms. Pauli emphasized that this treatment is mute about the claimed, the relation AEAt ~ h is analogousto therelation ApAq ~ h.
exact times of transition. Does this fact indicate a fundamental, irre- Yet the uncertainty relationsof time-energy andposition-momentum
ducible acausality, or is it merely a sign of the temporary incomplete- _ are not analogous. Thelatter is a straightforward deduction from the
ness of contemporary theoretical knowledge? Pauli (1926a, 11) refused mathematical formalism of the quantum theory; the former is not.
to take a stand: “This is very much debated,yetstill an unsolvedissue.” Heisenberg legitimated. the time-energy relation merely. by discuss-
Heisenberg (1927b),in responseto this challenge, proclaimed a “defini- ing several thought experiments—no wonder a complex controversy
tive” solution of the causality problem. _ developed around the validity and the meaning of the time-energy
Another fundamental problem, according to Pauli (1926a), was the
uncertainty formula, in contrast to the wide acceptance of the position-
duration of transition processes. Any attemptto introduce a detailed momentum relation.’ Heisenberg’s imprecise treatment of the time-
description of the transition process into quantum theory proved un- energy relation contained the seeds of future difficulties. In the
fruitful. Certain theoretical considerations lead to the conclusion that uncertainty paper, Heisenberg (1927b) presented what he called the
these processes must be very short compared with the lifetime of sta- “familiar” equation Et — tE = —if (as written there by Heisenberg).
tionary states. Nevertheless, Pauli continued to assume, in the spirit of
This time-energy formula is misleading; time, as many physicists
Bohr’s work, that stationary states are characterized by exact values of point out (see Aharonov and Bohm 1961), enters into the Schrédinger
energy, electric and magnetic moment, and so forth, and that these pa- equation as a parameter, not as an operator—the source of the basic dif-
rameters change discontinuously during sharp and precise moments of ference between the time-energy and position-momentum uncertainty
time. Pauli suggested that the precision limitof the time of transition is formulas. ,
perhaps of the same order of magnitudeasthe period of emitted light.
Pauli admitted that he could notoffer a more precise analysis. Dialogue with Dirac
Heisenberg immediately took up Pauli’s challenge to specify the pre-
cise limits on the determination of transition times: “According to the Heisenberg’s dialogue with Paul Dirac wasessentialto his formulation
physical interpretation aimedat here, the timeof transitions or ‘quan- andinterpretation of the uncertainty principle. Dirac’s work provided

14. Other examples Heisenberg referred to (both dueto Pauli) were of “weak quanti- 15. Here Heisenberg again refers the reader to Pauli’s discussion in his encyclopedia
article. °
zation,” when quantized periodic motionis interrupted by quantum jumps, and ofa ro-
tator in the form of a gear wheel. : 16. This controversy is discussed comprehensively in Jammer (1974).
 86 Chapter Four
The Dialogical Emergence of Heisenberg’s Uncertainty Paper
87
the mathematicaltools of transformation theory for the derivation ofthe
(Heisenberg to Pauli, 4 November 1926, PC). The problem
uncertainty principle. His achievement dramatically accelerated of interpret-
Hei- ing quantum mechanics thus became organically connected with
senberg’s conviction that the new quantum theory wasfinal, or com- that
of deciphering the basic commutation relation.
plete. This conviction(orat least strongly held assumption) wascrucial
In contrast to the authors of matrix mechanics, Dirac did not
for Heisenberg’s reinterpretation of experience (his definitions consider
of the noncommutativity a sign thata visualizable, geometrical
position and the path of a particle in operational terms). Dirac’s pow- interpretation
erful generalization of Heisenberg’s fluctuation paper transformedHe was fundamentally impossible in the quantum domain.In Dirac’s view,
i- noncommutativity was not counterintuitive, because it could be appre-
senberg’s preference for discontinuity into a passionate commitment.
hended in geometrical forms. Dirac was familiar with “non-Pascalian
Dirac’s stand regarding the essentially nonstatistical nature of quantu ”
m geometries, in which the “coordinates” of a point did not commute
mechanics, and his claim thatstatisticsis merely introduced by experi-
(Darrigol 1992a; Mehra and Rechenberg 1982). He consequently be-
ments, opened the wayfor Heisenberg’s legitimation and interpretation
lieved that the abstract relations between q-numbers, including the
of quantum mechanics through analysis of thought experiments. Di- ba-
sic commutation formula, could be given a geometric interpretation
tac’s impact on Heisenberg wasimportant inother wavsas well:
it re- (though not an interpretation continuous in space-time). Dirac made
inforced his kinematic explorations, andit stimulated further his search
noncommiutativity a central partof his version of quantum theory.
for physical insight through the analysis of Schrédinger’s wave func- Heisenberg adopted Dirac’s pointof view. The geometric (kinematic)
tion.-Dirac’s work strengthened Heisenberg’s growing preferencefor -
interpretation of the relation pq -- gp = (h/2mi)1 is the centerpiece of
.the abstract matrix formalism and Unanschaulichkeit, despite
Heisen- Heisenberg’s uncertainty paper. The general interpretive problem in
berg’s proclamation of an “intuitive” interpretation.
the paper is stated alonglines laid down by Dirac: which c-numbers
(coordinates in regular space-time) correspond to q-numbers (general
Centrality ofNoncommutativity abstract operators) that satisfy the fundamental quantum equation?
Heisenberg’s analysis of actual and imaginary experiments in his un-
The most nonclassicalresultof the original matrix theory was the non- certainty paper was intended to demonstrate that experimentalsitua-
commutativity of “position” and “momentum”in the basic equation
tions are consistent with the basic commutation relation of quantum
“Pq ~ gp = (h/2mi)1. Heisenbergatfirst regarded this noncommutati
v- mechanics.
ity as a hindrance or defect that might be eliminated as the theory de-
veloped further. For Diracthis noncommutativity was not a hindran Finality and Discontinuity
ce
but a sign of a fundamental advance. As Dirac explained vearslater:
In a letter written in the fall of 1926, Heisenberg informed Pauli of
someone(like Heisenberg) who takes a very daring and unconve
n- Dirac’s “extremely broad” generalization of Heisenberg’s own fluctua-
tional step can only advance upto a certain point—a natural anxiety
tion paper. Heisenberg presented Dirac’s reasoning to Pauli as follows:
(about being wrong) prevents the author from going further. Others
The general dynamical problem wasformulated by Dirac in fundamen-
therefore step in and carry on (interview with Dirac, AHQP).
tally probabilistic terms. If p and q are canonical variables, what can
Dirac’s version of quantum mechanics was built on symbols, g-
one say physically about a function f(p, q)? If gq is taken to be a specific
_ numbers, by analogy with the Poisson brackets of classical mechanics.
c-number, say, q = 10, classically one can calculate f(p, 10). In quantum
Dirac therefore viewed the noncommutativity in the basic relation as be-
mechanics this cannot be done—one can only specify the range of p
ing anatural counterpartof the noncommutativityin the Poisson bracket
for which f lies between the c-numbers f and f + df. Dirac succeeded
algebra. Rather than shying away from noncommutativity, he conceived
in defining the probability function in the form of a general matrix S,
quantum mechanics as the “general theory of all quantities that do not
which contained all the physically meaningful statements that could be
satisfy the commutative law of multiplication” (Dirac 1929, 716).”
made in quantum mechanics at the time, including statements about
Impressed by Dirac’s transformation theory, Heisenberg
arrived at Born’s collision processes and Jordan's canonical transformations. Hei-
the same pointof view: every mathematical scheme thatfulfills the
ba- senberg concluded enthusiastically, “I hold Dirac’s work to be an ex-
sic relation pq — qp = (h/2zi)1 is correct and physically meaning
ful traordinary advance” (Heisenberg to Pauli, 23 November 1926, PC ds
Dirac’s generalized version of quantum mechanics provided clear
17. Mydescription of the Heisenberg-Dirac dialogue relies heavily guidelines for moving from onerepresentation in terms of canonical
on the interpreta-
tion of Dirac’s workin the studies by Darrigol (1992a) and Kragh (1990).
variables to another. Incorporating work by Pauli and Schrédinger,
88 Chapter Four The Dialogical Emergence of Heisenberg’s Uncertainty Paper 89

Dirac demonstrated that the wave function % could be viewed as a Wesee thatinitially Heisenberg struggled with physical interpreta-
transformation from a scheme in which position is diagonal to one in tion in terms of a geometric analysis of concepts, and not in (opera-
which energy is diagonal.'* Because of the freedom contained in the tional) termsof the process of measurement. This approach did not lead
abstract conception of q-numbers,Dirac did not have to confine himself, far: “When wetalk about position or velocity,” he wrote, “we always
as the matrix theorists initially had, to matrix schemes in which the en- neéd wordsthat are obviously not definedatall in this discontinuous
ergy matrix is diagonal: This freedom provedpotent, especially in Di- world” (Heisenberg to Pauli, 23 November 1926, PC).
racs able hands (Kragh 1990; Jammer 1966; Darrigol 1992a). But Heisenberg needed a physical interpretation of the microdomain
Heisenberg elaborated on this feature of Dirac’s theory in his uncer- that used familiar kinematic concepts, both becauseof the theoretical
tainty paper, demonstrating how transformation theory supports the advancesof Pauli, Jordan, and Dirac and because ofSchrédinger’s con-
uncertainty relations and enhancesthe superiority of the matrix point tention that matrix mechanics was basically. nonintuitive. Pauli pro--
of view over all others (see chapter 5). For Dirac the existence of a videdthefirst clues: the position of a particle should be considered ob-
general, elegant, powerful set of transformations in quantum theory servable in principle, momentum and position cannotbe determined
suggested that his version of quantum mechanics compared favorably simultaneously, the path of a particle should be described in statistical
with the most developed parts of physics—Hamiltonian mechanics terms. Heisenberg exploited these insights fully in the uncertainty pa-
and relativity theory (Darrigol 1992a, 348). Transformation theory gave per. Yet Pauli did not suggest, as far as the available evidencereveals,
Dirac strong incentive to consider quantum mechanics complete. thatthe particle-kinematic interpretation of quantum physics be elabo-
Heisenberg became an eager advocate of the completeness, or final- rated in operational terms. This decisive step was Heisenberg’s, and it
ity, of quantum mechanics. He also madea crucial step toward discon- was taken, as I will shortly argue, underthe influence of Dirac’s and
tinuity: “I hold it now more than everas completely out of the ques- Jordan’s work. . .
tion that the world is continuous” (Heisenberg to Pauli, 23 November Both discontinuity and finality were crucial in Heisenberg’s rea-
1926, PC). His enthusiasmsfor thefinality and for the discontinuity of " soning in the uncertainty paper. Discontinuity entered fundamentally
quantum mechanics reinforced each other. Thefact that Dirac’s power- - (though incorrectly) into his analysis of the “uncontrollable discon-
ful generalization was based on Heisenberg’s discontinuity considera- tinuous” Compton recoil (disturbance) during an idealized position
tions further motivated, at the emotionallevel, Heisenberg’s embrace méasurement. From this point on, Heisenberg considered discontinuity
both of discontinuity and completeness. the mostessential feature of quantum mechanics and claimed thatits
Yet atfirst this reinforced belief indiscontinuity did not make matters importance could not be overemphasized. This emotional attachment
easy. Heisenberg did not see how one could possibly obtain a physical to discontinuity, canonized in the misleading notion of disturbance,
interpretation in terms of familiar concepts in the quantum domain. In is probably the reason for Heisenberg’s lifelong’use of disturbance im-
this discontinuous world waves and particles seemed completely alien: agery. The appealing imagery of the disturbance concept, backed by
“But as soon as it [the world] is discontinuous,in all of our words that Heisenberg’s authority, would confuse many students and interpreters
weuse for the descriptionof a fact, there are too many c-numbers. What of the philosophy of quantum theory.”° _
the word ‘wave’ or ‘corpuscle’ means, one no longer knows” (Heisen- Belief in the finality of quantum mechanics, strengthenedin dialogue
berg to Pauli, 23 November 1926, PC). with Dirac, shaped Heisenberg’s interpretative efforts. Withoutthis be-_
Heisenberg’s only consolation at the time washis intuitive “under- lief, on encountering a discrepancy between “nature”(the continuous
standing” of why p and q cannot haveexact values simultaneously. If path ofa particle) and “formalism” (an inability to describe space-time
space-time is discontinuous, the velocity at a definite point cannot have trajectories), Heisenberg would seek to improve the formalism rather
any meaning, becausein order to.define velocity at a point, one needs a than to reinterpret nature. As I will argue, Heisenberg’s opinion that
second point infinitely close to thefirst: impossible in a discontinuous quantum mechanics was complete determined the direction of his con-
world (Heisenberg to Pauli, 23 November 1926, PC Ve cludinginterpretive steps in the dialogue with Jordan.

18. Jordan developed a similar theory, in which he studied transformations from one 20. Heisenberg continued to rely on disturbance imagery after Bohr’s explicit critique
canonicalpair of variables to anotherpair ina givenrepresentation.
of this notion. Let us mention that Bohr, despite his critique of the disturbance notion,
19. “Dass die Welt kontinuierlichsei, halte ich mehr denn je fir ganzlich indiskutabel.” nevertheless invoked it because it had “intuitive” appeal(see chapter 12).
 90 Chapter Four
The Dialogical Emergence of Heisenberg’s Uncertainty Paper 91
However, this stand (both by Dirac and by Heisenberg) was prema-
ture; formidable problems remained to be solved (Darrigol 1992a, view,” very important. By analogy with relativity, itimplied that some
338). variables—position, for example—are frame dependent(in the
For Dirac,the criterion of “mathematical beauty” reinforced his belief posi-
in the finality of quantum mechanics. For Heisenberg, whose tion representation, positions have an exact value, while in the momen-
philoso- tum representation, they are completely undefined). The
phy was an opportunistic mixture of mathematical Gottingen and need for an

has
physical Copenhagen, the embrace of finality and discontinuity was analysis of observational possibilities was then a natural conclusion.
spurred byhis desire to rout Schrédinger’s waves once andfo¢all. Dirac’s somewhatlater statement sumsupthis trend of thought in a
strong form: “This state ofaffairs . . . [implies] an increasing recognition
Appeal to Experiment- of the part played by the observer in himself introducing the regulari-
Heisenberg’sinitial deliberations on discontinuity did not contain any ties that appear in his observations” (1930,v). i
analysis of experiments (actualor idealized) as a possible solution to
Heisenberg’s reasoning in the uncertainty paperechoed Dirac’s delib-
the interpretive problem. Nor did Heisenberg, who wasat the time with erations. In the abstract of the paper Heisenbergstated that the uncer-
Dirac in Copenhagen, mention the need to analyze experiments while tainty formula “is the real basis for the occurrenceofstatistical relations
in quantum mechanics” (1927b, 62). Heisenberg transformed Dirac’s
outlining to Pauli the essence of Dirac’s advances. Nevertheless, quan-
question “Howis statistics introduced by experiments?” into “How is
tum theory as.conceived by Dirac contained a puzzle: it was built ona
uncertainty introduced by experiments?” 2 His answer inaugurated the
statistical foundation, even as most recentexperiments—Bothe-Geiger;
thorny “disturbance” explanation: In experiments used for the opera-
for example—seemedto rule against fundamental indeterminism: This
tional definition of concepts, microscopic particles “suffer an indeter-
- . conflict had to be faced.
minacy introduced purely by the observational procedure we use when we°
In the fundamental paper that presented the essentials of transfor-
ask of them simultaneous determination of two canonically conjugate
mation theory, Dirac (1927) also tackled the question of physical inter-
quantities” (Heisenberg 1927b, 68, myitalics). Heisenberg identified, in
pretation in general, and the causality issue in particular. Dirac, unlike
the disturbance concept, the principal distinction between classical and
Born, did not consider quantum theory essentially statistical. The sta-
quantum mechanics. In classical mechanics(as in quantum mechanics)
", tistical element, Dirac argued, is only introduced when experiments
one can only give the probability of a definite position of an electron if
"are performed and questions are formulated using the information ob-
one does not know the phases of motion, yet in the classical case one
tained: “The notion of probabilities does not enter into the ultimate
can always conceive of determining phasesprecisely, through suitably
description of mechanical processes; only when oneis given somein-
chosen experiments. In quantum mechanics, however, such determi-
formation that involves a probability . . . can one deduce results that
nation “is impossible, because every experimentfor determination of
involve probabilities” (Dirac 1927, 641). In letters to Pauli and Jordan,
phase perturbs or changes the atom” (myitalics). In a definite stationary
as well as in his uncertainty paper, Heisenberg sided with Dirac on
state, the phasesare in principle undetermined, as follows from the ba-
this issue (Heisenberg to Pauli, 23 February 1927, PC).?! Mightnot anal-
sic equations Et — tE = —ih and Jw — w] = —ih (Heisenberg 1927b,66).
ysis of experiments prove helpful for understanding other nonclassical
peculiarities of quantum theory? Heisenberg could naturally have in-
quired. Dialogue with Jordan
Dirac’s attitude toward experiments was influenced by his opinion:
Heisenberg’s dialogue with Pascual Jordan wascrucial in the last phase
that there is an analogy between the generalized version of quantum
theory andrelativity theory (the freedom to move from one scheme to of his search for an adequateinterpretation of quantum mechanics.?? It
greatly reinforced Heisenberg’s growing tendency to bring the analysis
another). This analogy suggested the introduction of the notion ofex-
of experiments into the interpretive picture. In direct response to Jor-
periment, or even of an observer, into the interpretive search. As Dar-
dan, Heisenberg formed a structure of argumentation that unified pre-
tigol (1992a) points out, Dirac considered the existence of the group
of quantum mechanical transformations, relating different “points of viously fragmented insights. Heisenberg, as his correspondence with

21. “As Dirac does, so I believe that. ; - all statistics is only introduced through 22. “Onecan say, if one will, with Dirac, that the statistics are brought in by our
our ex-
experiments” (Heisenberg to Pauli, 23 February 1927, PC): periments” (Heisenberg 1927b,66).
23. My description of the Heisenberg-Jordan dialogue followsBeller (1985) closely.
 92 Chapter Four The Dialogical Emergence of Heisenberg’s Uncertainty Paper 93

Pauli reveals, struggled with the issues raised by Jordan (Heisenberg to during measurement, and consequently, the disturbances could be
Pauli, 5 February 1927, PC). But so great was the inspiration that within eliminated (for details, see Beller 1988). Yet there always remained a
two weeks, he had a response:the core of the uncertainty paper (Hei- conspicuousdisturbance of the zero point.
senberg to Pauli, 23 February 1927, PC). Ising (1926) analyzed Moll and Burger’s data and demonstratedthat
theperpetual fluctuations of the zero point were simply Brownian fluc-
Jordan's Discussion of the Causality Issue tuations. Ising’s paper, which implied limits on the accuracy ofall
measurements, did not cause any conceptual stir. The widespread as-..
Jordan’s paper (1927b) was devoted to exploring the relation between sumption wasthat by reducing the temperature of the measuring de- |
the foundations of quantum mechanics and the problem of causality. vice one could reduce the amplitudeof the fluctuations, which should’
Jordan stated that causality is not an a priori necessity of thought but approach zero as the temperature approaches0°K. Yet Zernike (1926)
rather a question that should besettled by experiment. Physicists can- soon showedthat evenin this case the fluctuations would persist. Al-
notbe satisfied by an approximate metaphysical determinism:to arrive though the oscillations of a freely swinging galvanometer can-be re-
at a definition of determinism, the physicist must specify conditions duced by lowering the temperature, there are fluctuations in the charge
under which its existence can be verified. The definition of determi- of the external circuit to which the galvanometer is connected. A gal-
nism must therefore change in accordance with theories and experi- vanometer will exhibit these fluctuations even if its own temperature
mental methods. Determinismin classical physics means both complete is O°K. oes co : os oo
TisAGELWILLREAST) dee

knowledge ofinitial conditions (in biology, for example, such knowl- Whatis the connection between-quantum indeterminacy and the lim-
edgeis unattainable) and the special characterof physical laws (second- its of actual measurement? “When we rememberthatthis is the case
order partial differential equations in a four-dimensional manifold, with all our apparatus, andthatit all ‘rattles about’ in this way, we may: °
with one dimension, time, being imaginary). In the classical case, the be temptedto think that the experimentalist is quite as incapable ofobsero-
present will uniquely determine the future course of events, because ing elementary processes as the quantum theorist is of predicting them” ar-
in the classical world, quantities are continuously propagated through gued Jordan (1927b, 568, my italics). Yet one could avoid these experi-
space and all physical motions are continuous. mental uncertainties, he continued, by making experimentsat absolute
In the quantum domain one must be prepared to sacrifice this kind zéro or, moré comfortably, by working with particles with a vast store
of determinism becauseofthe existence of elementary discontinuities — of energy. In such cases, one could observe in Wilson's cloud chamber
quantum jumps. Quantum lawsare indeed at presentstatistical; in the trajectory of a single particle and determine exactly the moment
general they say nothing abouta single atom butyield only mean prop- whenthetrajectory ended in a quantum jump.
erties of an assembly of similar atoms. If, however, one directs one’s Thus, Jordan stated, the time of a quantum jumpisin certain cases
attention not to discontinuities but to probabilities that evolve accord- exactly determined. How doesthis fact compare with available theory?
ing to Schrédinger’s differential equations, one can conclude that de- Because quantum theory is statistical, one might concludethat it can
terminism (for these probabilities) holds by analogy with the classical only providethe probability of a quantum jump, andthattherefore the
principles (Jordan 1927b). time of transition is undetermined. Yet this conclusion doesnot follow
Whatis the connection betweenthestatistical character of quantum from the foundations of the theory; it was an additional hypothesis of
laws and experiments? Initially, one is temptedto conclude that experi- the Bohr-Kramers-Slater theory, which was disproved by experiment.
ments would give nothing but averagevalues. This conclusion could be In this respect, it is significant that Pauli’s assumptions say nothing
basedon the results of Frits Zernike (1926) and Gustaf Ising (1926), who about the probabilityof a transition (theyrefer to the probability that a
demonstrated the existence of impassable limits of accuracy in mea- system will be at a given place in configuration space). In orderto deal
surements, due to Brownian movement.Ising’s and Zernike’s work was with this issue, one has to reduce the probabilities of quantum theory -
triggered by the attempt of two experimentalists, Erich Moll and Her- to elementary probabilities. “Only then can wesay that wereally un-
man Burger (1925), to improve the accuracy and sensitivity of measur- derstand the laws; and only then can wetell under what conditions the
ing devices. By constructing a “thermalrelay” device they achieved time oftransition is determined. Only then can we know exactly what
an unprecedented improvement in the sensitivity of galvanometers. is causally determined, and whatis left to chance.” The question of de-
This device allowed the detection of the origin of various disturbances terminism in the quantum domain remains open (Jordan 1927b, 569).
 94 Chapter Four
The Dialogical Emergence of Heisenberg’s Uncertainty Paper 95
Heisenberg’s Response to Jordan's Discussion
5‘. From points 3’, 4, and 4’ it follows that one can give a conclusiv
? e
It is convenient to summarize Jordan's theses as items 1 through 5 be- answer to the question of determinism:there is no determinism in
the
low, and then to follow with Heisenberg’s direct responsesto them, quantum domain.
ignated items 1' through 5’, in a tacit dialogue with Jordan des-
j Roints 1’ through 5’ are the central theses of Heisenberg’s uncertainty
paper. Heisenberg was familiar with Jordan’s paper, having read it with
1. There are experimentalsituations that cannotbe describ
ed by the grossen Genuss (Genuss means both “joy” and “profit’). Heisenb
formalism of quantum mechanics(the trajectoryofa‘particle, the erg
de- mentioned Jordan’s workin connection with his attemptto understand
terminate time of a quantum jump in certain cases). Thus
quantum the logical foundations of the pq — qprelation, and inspired by Jordan’s
theoryis as yet incomplete.
explorations, he began contemplating whatthe position of an electron
2. One maythinkinitially that experiments,like theory, contain essen- means.In his paper, Jordan had inquired about the “probability offind-
tial indeterminacies due to Brownian motion. Yet these experi
mental
ing an electron in a certain place.” Heisenberg found this expression
indeterminacies can be avoided by choosing experimental
conditions unrigorous. On the basis of his previous stand that an electron’s posi-
properly. tion is unobservable, Heisenberg did not consider “the place of an elec-
tron” to be a defined concept: “What; for example, does ‘probability
'3. Quantum theoryis essentiallya statistical theory. Yet it has not been for an electron to be in a definite point’ mean when the concept‘posi-
reduced to independentprobabilities. Therefore, quantum theory is as
tion of an electron’ is not appropriately defined?” (Heisenberg to Pauli,
yet incomplete. , a? 5 February 1927, PC). Just two weekslater, Heisenberg answered this
4. The question of determinism is one of both the completeness with question in anotherletter to Pauli, in which helaid outall the essentials
which the initial conditions are known andthe special form of the of his uncertainty paper (Heisenberg to Pauli, 23 February 1927,PC). In
physical laws. : : the spirit of points 1’ through 5’, the question aboutan electron’s posi-
tion was answered in operational terms.
5. Because of point 3, one cannotyet give a conclusive answer to the Heisenberg’s discussion of the limits of measurement was directly
question of determinism in the quantum domain. suggested by Jordan’s paper. Heisenberg sought quantum mechanical
We know from Heisenberg’s correspondence and his recollections limits on measurements by analogy with classicallimits in thermody-
that he regarded quantum theory atthat timeas final, and its mathe- namicterms. Bohr, familiar with Heisenberg’sefforts, was later explicit
matical formalism as complete. Thus Heisenberg’s responses to points on this point: “It is true, as emphasized by Heisenberg, that an instruc-
1 through 5 run asfollows: tive analogy to the quantum theoretical point of view is obtained by
comparing the uncertainty in the observations of atomistic (micro-
1’. Because quantum theory is already complete, there can be no ex- scopic) phenomena with the uncertainty inherently contained in any
perimentalsituations that cannot be described byits formalism.If observation, due to imperfect measurements, as considered in the or-
one
finds such situations (such as the path of a particle), they must be re- dinary description” (1927c, 123). Perhaps it is worthwhile to add here
interpretedso as to fit the formalism. that Zernike’s paper (1926) aboutlimits on the accuracy of measure-
ments was published in the sameissue of Zeitschrift fiir Physik where
2'. One must be able to demonstrate the existence of unavoidable
ex- Jordan’s transformation theory was elaborated (Jordan 1927f). Heisen-
perimental indeterminacies, in order to have agreement
with theoreti- berg, of course, read this paper of Jordan’s (as he wrote to Pauli, 5 Feb-
cal indeterminacies.
ruary 1927, PC). It is unlikely that he did notat least glance through
3’. Quantum theory is complete. Therefore,it is not essentially a sta- Zernike’s work, after it had been called to his attention by Jordan’s dis-
tistical theory. , cussion. Thusthat Ising and Zernike had an impact on Heisenberg’s
thoughtis highly plausible. The realization that there are inevitable
4’. Because quantum theory is not essentially statistical, limits on the accuracy of real measurements provided Heisenberg with
the ques-
tion of indeterminism reducesto the question of the completeness the crucial clue: he would pursuean intuitive interpretation of quan-
of
knowledgeof initial conditions. Because of point2’, this knowled
geis tum mechanics through an analysis of the limits of measurementin his
essentially incomplete.
thought experiments.
 96 Chapter Four The Dialogical Emergence of Heisenberg’s Uncertainty Paper 97

Dialogues with “Lesser” Scientists interference experiments, where no temporal componentis involved,
and experiments on the measurementofthe velocity of light, in which
The names Zernike and Ising do not appear in the usual historical temporal magnitudes are associated not with atomic systems but with
works dealing with the genesis of Heisenberg’s uncertainty principle.
macroscopic bodies, such as planets or revolving mirrors. It is surely a
Nor do we find the names Duane, Campbell, or Sentfleben.* Yet all hasty step “to say that we have observedin ‘principle’ the frequency
these scienSsts had prominent places within the dialogical web that belongingto the latter when we have actually measureda frequency be-
formed as Heisenberg groped toward the formulation of the uncer-
longing to the former” (Campbell 1926, 1115). > ate
tainty principle. By burying such namesin theannals ofhistory, the Contrary to Heisenberg, Campbell(1926) suggested considering time _
usual accounts suppress the dialogical nature of reasoning and enhance
as a purelystatistical concept and, consequently, such notions as ve-~
the “hero worship” tradition. In “revolutionary” accounts the steps locity or frequency as applicable only tostatistical aggregates: Camp-
taken bythe central players seem more discontinuous (and irrational) bell supported his proposal with an analysis of measurementby radio-
than they actually were. In linear “rational reconstruction” accounts, active clocks.In the case ofthe radioactive clock,thereis no possibility
the complexity and ingenuity of eachscientist’s complex reasoning is of defining a time interval by observing a single atom—only a large
hardly apparent. A dialogical analysis of the achievements of our sci- collection of atoms allows measurementsof temporalintervals.
entific heroes puts into proper light their outstanding creativity. Such In his discussion Campbell raised a number of fundamental issues.
_an analysis also resurrects the “lesser scientists,” bringing back to life that wereclose to Heisenberg’s interpretive concerns:
the relevancy, ingenuity, and indispensability of their thought.
1. Campbell's analysis of the measurementof frequencies impliedthat
Dialogue with Campbell
such observables are theory Jaden. Such an analysis could have rein-
The revolutionary realization that time-space concepts do not apply . forced Einstein’s famous dictum that “it is the theory which decides
in quantum mechanics is usually ascribed to Bohr, and the ingenious what wecan observe,” which, according to Heisenberg’s recollections,
implementation of this realization in the mathematical formalism of was crucial to Heisenberg’s discovery of the uncertainty principle
quantum mechanics to Heisenberg. Yet many other scientists before (1971, 63).
"Bohr and Heisenberg contemplated the overthrow of mechanics and
" 2.’ Campbell's suggestion that timeis statistical in nature becamea fo-
the modification of space-time concepts: Planck, Poincaré, Richardson,
cus of Heisenberg’s deliberations about thestatistical nature of quan- _
Sommerfeld, Campbell, and Sentfleben, among others.
tum theory. Just a few months before his formulation of the uncer-
After the appearance of matrix mechanics, Norman Campbell (1926) . tainty principle, Heisenberg wrote to Pauli that perhaps “space and
repeated and extended his earlier call (Campbell 1921) for a modifica- timeare in reality only statistical concepts, such as temperature, pres-
tion of space-time concepts in the quantum domain. Campbells papers sure, etc, ina gas. I mean that the space-time concepts for an individual
are significant because they concentrate on the necessity of modifying corpuscle are meaningless” (Heisenberg to Pauli, 26 October 1926,
the conceptof time and redefining the notion of the path of an intra- PC). Heisenberg’s words are remarkably close to Campbell's assertion:
atomic electron. Both insights were used in Heisenberg’s uncertainty “The view I wantto suggest here is that time, like temperature, is a
paper. Campbell (1926) connected the existence of paradoxes within the purely statistical conception, having no meaning exceptas applied to
quantum description with the uncritical adoption of the time concept. statistical aggregates” (1926, 1107). Heisenberg eventually abandoned
in matrix mechanics. Although matrix mechanics arose from Heisen- this approach and beganto explore thelimits of applicability of space-
berg’s daring reinterpretation of the space container, Heisenberg as- time concepts for an individual atomic particle. Yet Campbell’s insis-
sumed, for no good reason, that the classical concept of time applied
tence that the analysis of space-time concepts and the problem of cau-
sality were closely related time and chanceare merely two aspects
withoutalteration to atomic systems. Such a conceptof time underlies,
of the same thing” and “time and chance are inevitably associated”
for example, Heisenberg’s statement that frequenciesof radiation belong (1926, 1107, 1108)—reverberates through Heisenberg’s uncertainty pa-
to entities observable in principle. But this statement is controversial: per. Campbell’s discussion could have suggested to Heisenberg the
the determination of frequencies demands two types of experiments — formulation of the time-energy uncertaintyrelation in probabilistic
terms. Campbell's discussion of the interrelation of space-time and
24. As far as I know, only Hendry (1984) noticed the connection between Sentfleben
’s causality wasalso a trigger-for Bohr’s discussion of the complemen-
work and Heisenberg’s.
tarity of space-time and causality (see chapter6).
 98 Chapter Four
The Dialogical Emergence of Heisenberg’s Uncertainty Paper 99
3. There is another point in Campbell's analysis that resonat
es in Hei- in whichtheaffixes are different and those in which they are the same,
senberg’s interpretive attempts. Campbell suggested
abandoning the
usual temporal concepts, yet he clearly realized that the concept butthese latter can be associated only with transitions within the same
of
time is deeply embeddedin our thought and language and stationary state” (1926, 1115). 7
that with-
outit “we could not even state adequately the problems wedesi
re to Campbell does not offer any concrete physical theory—only guide-

fue
solve” (Campbell 1926, 1110). Is it possible to proceed
in such a case? lines for the principles on which the future quantum theory should be
Yes, Campbell replied to his own question, but onlyif built. He apologized for “offering such a-collection of incomplete
one abandons and
formal consistency. By using temporal conceptions
freely in our con- incoherent suggestions” and then concluded prophetically: “If anyone
structive efforts while simultaneously denyingtheirvalidity, we can make any use of them, he will be muchbetterentitled than myself
might
eventually realize the direction of a future consistent
solution. to be regarded astheir author” (1926, 1117).
Howcan one explore Campbell’s suggestion that the concept Dialogue with Sentfleben
of time
does not apply to an individual system? Campbell's
answer is experi- While peripheral in the context of the issues discussed in the uncer-
--Mentoriented: “The value of these ideas would be proved
decisively if tainty paper, it is Heisenberg’s verdict on causality that attracted the
it were possible to produce a theory . . . involving no temporal con-
attention of philosophers and philosophically mindedscientists. Hei-
ceptions ... and at the sametime to show thatall the experiments
on senberg announcedthat his discussion established “thefinalfailure of
_ which the prevailing temporal conceptions are based can:be described
causality” (1927b, 83). He explained confidently: “In the strong formu-
in termsof Statistics’ (1926, 1110). As follows from Heisen
berg’s let- lation of the causal law, ‘If we know exactly thepresent, we can predict
ters to Pauli, Heisenberg repeatedly hit dead-ends when analyzing such
the future’ [it] is not the conclusion but rather the premise, which is
kinematic concepts as “velocity” in the “discontinuous world.
” Camp- false” (1927b, 83). Since we cannot knowthepresent precisely, we can-
bell’s suggestion that one could clarify the application of the concep
tof not predict the future with certainty.
time through an examination of experiments, without achieving prior
analytical clarification of the concept, was used in Heisenberg’s This solution was a novelone.In classical physics it was assumed that
uncer- in principle (though not in practice) one can know the presentexactly
tainty paper.
Another fundamental point, which was commonto Campbell's and (or at least Laplace’s demon can). Accordingto the philosopher Moritz
Heisenberg’s discussions, is the reinterpretation of the concept of Schlick, Heisenberg’s solution wasa great surprise for modernphiloso-
the phy “since even the mere possibility of such a solution had never been
‘orbit of an intraatomic electron. Both Campbell and Heisen
berg con- anticipated,” despite frequent discussion of this problem through the
ceived of such orbits as sequencesof discrete points. One of
the for- ages (quoted in Jammer 1974, 75).
midable problems Heisenberg struggled with was how to introd
uce Schlick’s assessment was wrong: H. A. Sentfleben, a physicist whose
space-time concepts foran individual particle without simult
aneously name does not appear in the usual accounts of the development of
resurrecting the continuousintraatomic path. While Heisen
berg’s solu- quantum physics, reached a similar conclusion four years before Hei-
tion goes beyond Campbell's, Campbell's deliberations nevertheless
of- senberg (Sentfleben 1923). Sentfleben argued thatin principle the pres-
fered a new way of conceiving the path of an electron as an irregu
lar ent cannotbe precisely defined; therefore, the future can only be known
series of discrete points. According to Campbell, a particle that
is mov- in statistical terms. Sentfleben claimed that space-time mustbe discon-
ing along an orbit is actually undergoingirregulartransitions,
the locus. tinuous because the size of Planck’s constanth sets the limits for the
of which is a closed curve. Campbell perceived “stati
onary states as precision with which space-time concepts can be defined. According
consisting of transitions, just as much asthe changes from one
station- to Sentfleben, one need not give up space-time concepts(as the origi-
ary state to another” (1926, 1114). While quantum jumpsrelate
to tran- nal matrix theory later did), but only determinethe limits of their ap-
sitions between states of different mean energy, Station
ary states are plicability.*
transitions between states of the same mean energy. Campbe
ll consid-
ered his suggestion to be in the spirit of Heisenberg’s, Born’s,
and Jor- 25. The realization that the introduction of h demands a modification of space-time
dan'smatrix theory: “Their analysis dependsonthe use of double concepts was understood as early as 1911 by Planck, whoatthefirst Solvay conference
-affix
quantities, such asy,,, , which are associated with transit called for a modification of classical mechanics. Planck argued thatin orderfor the energy
ions, and in the
formal treatment they make no essential distinction betwee of an oscillator to be an integral multiple of the energy elementhv, phase space must
n quantities consist of finite elementaryareas h. This issue is discussed in Jammer(1966, 52-56).
 The Dialogical Emergence of Heisenberg’s Uncertainty Paper 101
100 Chapter Four

Sentfleben’s thoughts on the modification of space-time were con- expressis verbis on the ground of newest de velopments” (quoted in Dar-
nected with his agenda to obtain a unified foundation for physics. rigol 1992a, 264). Both Heisenberg and Pauli realized the relevance of
Sentfleben,like Einstein before him, aimed at avoiding the dualism of such a division to their attempts to understand the kinematic meaning
discrete atomistic conceptions and continuousfield theoretical ones. of the quantum mechanical formalism (Heisenberg to Pauli, 15 Novem-
“Sentfleben identified the source of the contradictionsin the old quantum ber 1926, PC). In this regard, Heisenberg commentedin the uncertainty
theoryas an illegitimate extension of continuousfield theoretical no- paper: “Equation (1) [piqi ~ h] is a precise expression for the facts
tions into the discrete microdomain.In order to build a consistent mi- which oneearlier sought to describe by the division of phase space into
d
crophysics, one has to reinterpret the notions of space and time. Be- cells of magnitude h” (1927b, 65). But while Pauli and Dirac reasone
cause description in terms of space-time notions is always connected in terms ofthe precise determination of one conjugate variable at the
with phenomena in which h plays a substantial role, Sentfleben intro- price of the complete indeterminacy of another, Heisenberg formulated
ducedthe following principle: ‘The possibility to describe space-time the problem in termsof the degree of precision of ajointspecification
events with arbitrary precision is limited in principle, where Planck’s of p and q (in termsof the size of cells h in phase space). It is this psy-
constant h is the factor which determinesthe limits of precision by chological leap from the world partly known into the world thatis im-
from
which a coincidence can be set between a matter-point and a space-time precise in principle that distinguishes Heisenberg’s deliberations
point” (1923, 131). those of Pauli and of Dirac. In this respect, there is a strong similar-
According to Sentfleben (1923), the world is discontinuous in prin- ity between Heisenberg’s and Sentfleben’s approaches. Throughoutthe.
ciple (let us recall how Heisenberg struggled to define kinematic con- uncertainty paper, Heisenberg, as did Sentfleben, used the word Un-
or
cepts in-such a discontinuous world). For small distances and times, genauigkeit (imprecision) rather than Unbestimmtheit (indeterminacy)
clocks’ and rods become useless because of their own characteristic Unsicherheit (uncertainty). After Heisenberg formulated the problem
for a
oscillations. Anticipating Heisenberg’s analysis, Sentfleben argued that along theselines, the uncertainty formula immediately followed:
microparticle in a cell of size h, the joint specific ation of p and q must
the problem ofthe applicability of space-time notionsis intimately con-
nected with the analysis of limits on measurements. In particular, due obey ApAq ~ h. Knowing the desired conclusion beforehand, Heisen-
ele-
to the molecular fluctuations of measuring clocks, the time of an event | ~ berg could easily deduce it from the mathematical formalism by
mentary useof transfo rmation theory.
cannot be defined with unlimited precision.
A new discontinuous quantum theory would clearly be opposed to
field theoretical macroscopic theories, in which strong determinism
holds, argued Sentfleben. This new theory mustbe formulatedin statis-
tical terms. Because Planck’s constanth limits in principle the possibility
of describing a space-time process with arbitrary accuracy, Sentfleben
concluded that the usual formulation of the law of causality (“Given a
space-time situation A, we can determinea later situation B”) is inade-
quate: “We simply cannot have a precisely defined situation A” (quoted
in Hendry 1984, 33). The similarity with Heisenberg’s discussionin the
uncertainty paperis striking.
The idea that h determines the structure of space-time is fully ex-
pressedin the division of phase space into finite elementary areas of
size h (cells). The division of phase space into quantumcells of size h
was a commonprocedure, used in many quantum theoretical papersat
the time. Schrodinger understood the direct link between Heisenberg’s
uncertainty principle and Planck’s quantum cells. In a letter to Planck,
Schrodinger wrote: “The partition of phase space in cells which you
[Planck] introducedfor statistical purposes, already contained the fun-
damental uncertainty relation, which was only muchlater established
 CHAPTER 5
ES
The Polyphony of Heisenberg’s
Uncertainty Paper

He [Heisenberg] is very unphilosophical.

WolfgangPauli to Niels Bohr, 11 February 1924, PC, 143

The one who insists on never uttering an error must remainsilent.

Werner Heisenberg 1958, 86

Introduction

Heisenberg’s reasoning in the uncertainty paper emerged from a com-

plex network of dialogues, intellectually challenging and emotionally
charged. The gradual forming of preferences, of choosing one intel-
lectual option over another, of shaping opinions, occurred as insights
gained during many different dialogues coalesced. I arguedin the pre-
vious chapter that Heisenberg’s preference for acausality formed in
reaction to Schrédinger’s space-time causal program, andin dialogue
with Pauli’s deliberations about the problem of determining the times
of intraatomic transitions. Heisenberg’s elaboration of the causality
issue received further impetus from Campbell’s discussion of chance
and space-timein quantum theory, from Sentfleben’s discussion of cau-
sality, and from Jordan’s doubts about the finality (and therefore about
the inherently statistical character) of quantum theory. Heisenbere’s
opinion that the quantum mechanical formalism was final crystal-
lized gradually. Crucial in this respect was the perceived needto insu-
late matrix mechanics from Schrédinger’s competing attempts, and the
strong support from Dirac’s theoretical advances. Heisenberg’s decision
to undertake an analysis of thought experimentsas the focus ofhis in-
terpretive attempts emerged in the last stages of the creative process,
in dialogues with Dirac, Campbell, and Jordan. Not the magisterial
unfolding of a single argument, but the creative coalescence ofdiffer-
ent arguments, each reinforcing and illuminating the others, resulted
 104 Chapter Five
The Polyphonyof Heisenberg’s Uncertainty Paper 105

in Heisenberg’s monumental contribution to physics. In the process of

that wave packets disperse, “only because in recent times they seem to
Heisenberg's discovery, the communicative natureof thoughtwas fun-
be forgotten” (1927b, 73).
damental.
Heisenberg’s paperis not a systematic monologue, in which a reader
It would be surprising if the final product arising from such a plu-
follows the unfolding of a central argumentto its end. We will see that
tality of issues, challenges, and doubts were unambiguous and un-
Heisenberg stopped his deliberations abruptly in order to send the
“equivocal. It would be surprising if the polyphonyof the creative act
paper for publication.It is contingent, I will argue, which voices Hei-
could be systematized into a neat conceptual scheme. Heisenberg’s un-
senberg chose to emphasize and which voices he decided to withhold.
certainty papercontains in fact traces of past struggles, vacillations on
Someof these voices, suppressed at the last moment, reemergedin Hei-
central issues, subdued, but continuing doubt aboutthe fundamentals.
senberg’s future publications; other voices perhaps werelost... °
Conflicting voices permeate the text. And the equivocations and contra-
Lookingat Heisenberg’s paper in this way, we cannotsustain theil-
dictions of the uncertainty paper reverberate insubsequent writings by
Jusion that Heisenberg wrote the uncertainty paperin order to convey
Heisenberg and other Gottingen-Copenhagen physicists.
an unequivocal message, the meaning of which was clear to Heisenberg
The fundamentally new formula of quantum mechanics in the matrix
himself, and that therefore the job of historians and philosophersis to
formulationis the basic commutation relation pq — qp = (h/ 27ri)1. We
decipher Heisenberg’s original message. Heisenberg, who was waver-
sawthatinitially the authors of matrix mechanics deduced nothing in-
ing on manyissues, hopedthat the very act of writing, together with
teresting or physically meaningful from this formula:? outside a dia-
Pauli’s response, would help him to “get some sense of his own consid-
logical context formulas are mute. Heisenberg’s reasoning in the uncer-
erations” (Heisenberg to Pauli,27 February.1927, PC). These considera-
tainty paper wasaimed at connecting the uncertaintyrelations with the _ tions Heisenberglaid outin his letter to Pauli, and the text of this letter
commutation formula—yet how exactly they are related remained un-
wastaken almostverbatim into the published version. Pauli’s response
clear. Heisenberg was “seeing double” on issues of Anschaulichkeit, on
(whichis lost) seemsto have'servedat this stage more as reinforcement
the role of classical concepts in the quantum domain, and on the causal- than asclarification.
ityissue. He was even unsure about whatit meant to providean inter-
The plurality of voices in a scientific paper should not be confused
pretation, vacillating between an operational and a modeltheoretical
. with stable elements in the Gestalt model of creativity. According to
approach.It wasinthis context that Heisenberg introduced the thorny Wertheimer (1959), the creative act consists in the recognition of a
disturbance problem, and the ensuing confusion between inevitability
restructuring idea, which resolves the tension that exists between sepa-
and consistency that plague the Copenhagen philosophy. Similarly, the rate facts, or elements. The resulting newGestalt eliminates contradic-
contradictory voices of positivism (operationalism), model theoretical
tions, achieves unity, and provides insight. Using this model, Wert-
realism (the invariant features of a successful scientific theory refer to
heimer, and after him Miller (1984), described Einstein’s discovery of
genuine aspectsof reality), and conventionalism (physicists can freely the special theory ofrelativity. From the dialogical perspective, there
choosethebasic axiomsof a theory, worrying only aboutits consistency
simply are nofinal, stable elements, or facts— everything can be ques-
and empirical adequacy)are all present in Heisenberg’s paper.
tioned and doubted. Noris there total, final resolution of tension—
The voices in Heisenberg’s paper are not neutral or disembodied.
conflicting voices coexist, or are temporarily put aside. Rarely are they
Theyretain traces of the high emotionsthatcharacterized Heisenberg’s
completely extinguished.
confrontation with the wave theoretical approach. We hear triumph:af-.
A paradoxical tension exists between the opennessof a scientific text,
ter Heisenberg’s own ingeniousinterpretation matrix mechanics should
addressed to the future, andits solid roots in the past. This tensionis
no longer be considered nonintuitive, refuting Schrédinger’s accusa-
connected with another tension—that between innovation and conti-
tions (Heisenberg 1927b, 82). We hear arrogant but good-temperedsar-
nuity. Heisenberg endedhis letter to Pauli with an apology, and an
casm: Heisenberg reminded Schrédinger of the simple consideration
anxiety, that all his deliberations about uncertainty contained nothing
1, See also Holton (1991) on conflicting voices in Einstein’s early work in physics. new—they were merely an alter Schnee (old snow; Heisenberg to Pauli,
2. Except declarations about the symbolic and nonintuitive character of
the matrix 27 February 1927, PC). Almostall of us are familiar with suchfears.
theory. They are connected not only with natural doubts about the worth of
3. “Seeing double” is Shapin and Schaffer’s (1985) term, relating to Hobbes’s
phi- our contributionsbut also with an inner knowledgeoftheir dialogical
losophy. ,
connection with the work of others. When we are awarethat our ideas
 106 Chapter Five
The Polyphony of Heisenberg’s Uncertainty Paper 107
are rejoinders to those of others, how
can we be sure that we not merely
repeat but creatively respond? This is theory. This interpretation contains a quantization algorithm for any
the reason, I suggest, for such
frequent vacillation in the evaluation of our given observable (it determines what measurement
own work between two results, or values,
extremes—“deeply profound” and “utterly trivia are possible) anda statistical algorithm (it gives the probability of
l.” a
The dialogical nature of creativity explains possible measurementresult). Accordingto this approach,the quantum
why reinforcement, sup-
port from anothervoice, is so important. The more mechanical formalism is indeed no rhore than toolfor the description
novel the idea, the and prediction of measurementresults, as Bohrclaimed.
stronger the need for reinforcement. For how
can we know, except Yet the minimal instrumentalist interpretation serves more as
through validation by another's voice, whether a ref-
our ideasare daring or uge whenoneis pressed by paradoxes and contradictions (especial
reckless, profound or absurd? oO ly
in the case of the Copenhagen interpretation) than a consistent, com-
The Polyphonyof the Notionof Interpre mitted interpretive stand. Most physicists, Bohr and Heisenberg in-
tation cluded, wanted more: some feeling of understanding, ofilluminating,
Heisenberg’s uncertainty paper was aimed of explaining the kind of world the quantum formalism describes. The
at providing a satisfactory
interpretation of the new quantum theory. need for this kind of metaphysical grasp is not merely psychologi-
Today there are manyin- cal but social as well—the power of a successful explanation and the
terpretations of quantum theory —the Copenhagen,the
many worlds powerofthe effective legitimation and dissemination of a theory are
(Everett 1957; De Witt and Graham 1973), the realistic statis
tical (Balien- connected. This is one of the reasons that Bohr, despite his minimalin-
tine 1970), the (nonlocal):-hidden variables (Bohm
‘

1952; Vigier 1982), the strumentalist pronouncements, nevertheless invoked his principle of
modal (van Fraassen 1981, 1991; Kochen 1985;
Dieks 1994; Bub 1992; complementarity to “explain” the quantum mystery.
Healey. 1989), the quantum logical (Finkelstein
1965; Putnam 1968; One need notbe anaive realist to appreciate the need for an inter-
Friedman and Putnam 1978).* Yet there is no agreemen
t on the basic pretation moresatisfying than the minimal instrumentalist stand. Van
question: what does it meanto interpret a mathematic
al-physical the- Fraassen formulated the issue of interpretation in the following way:
ory? Nor is there an unequivocal answer
to another basic question: “What would it be like for this theory to be true, and how could the
whatis a theory?
Thelegacy of logical positivism implies that a physi world possibly be the waythis theory saysit is?” (van Fraassen 1991,
cal theory is a 242). Heisenberg had his own succinct definition of interpretation: to
partially
interpreted formal system consisting of an abstr
act formalism understand a theory meansdaraus klug zu werden(to get clever)—to get
Fand set of rules, or bridge principles, or corre
spondencerules, R that senseout of a theory, to get a feeling of enlightenment and insight.
connect the nonlogical terms of the formalism with
observable phe- Heisenberg’s definition shows why the question of interpretation is
nomena.° This notion of a scientific theory survives
in what Redhead a controversial one—in which circumstances onegets a feeling of un-
(1987)called the “minimal instrumentalist” interpreta
tion of quantum derstanding is personally, professionally, and culturally conditioned.
4. “It sometimes seemsas if there are as manydifferen Those scientists who feel at ease with mathematical formal structures
t attempts [to understand quan-
tum mechanics] as there are people who have seriously
madethe attempt” (Healey1989, often gain enlightenment from the analysis of a mathematical formal-
2). In fact, it is difficult eventolist all the exciting interpr ism. Often scientists are willing to make bold ontological predictions
etive work being done nowadays
as an alternative to the Copenhagen interpretation
Broglie alternatives edited by Cushing, Fine, and
. The recent volume on Bohmian—de relying on the features of mathematical structures. Physicists espe-
Goldstein (1996) includes such notable: cially are inclined to ascribe realistic meaningto the invariant features
examples as Diirr, Goldstein, and Zanghi (1996) and
Valentini(1996). For an overview of
this interpretive work, see Redhead (1987) and
Cushing (1996).
of a theory, as is demonstrated by the writings of Einstein, Dirac, and
5. There is no need to rehearse here numerous Born. And indeed mostrecent attempts to interpret quantum mechan-
objections to this approach,the dif-
ficulty of distinguishing between theoretical ang ics rely on the mathematical formalism in a significant way. Those who
observational terms being among the
most prominent. Sufficeit to say thatall historically
ing with Kuhn and Feyerabend, sprang from criti
sensitive philosophyofscience,start- shy away from mathematical formalism try to gain enlightenmentby
que ofthis positivist model for a sci- linguistic ingenuity. As medieval theologians, such as Maimonides and
entific theory. It is worthwhile perhapsto addtha
tin the case of quantum theory, even
the notion of an abstract formalism is problematize Aquinas, ascribed apparent contradictions in the discourse about God’s
d: philosophers disagree about the
best wayto formalize quantum theory. There is no
unique formal system whose seman-
tics is open (Healey 1989,
5). 6. Leverrier’s prediction of Uranus, Dirac’s of the positron, and Gell-Mann’
. s of the
quark are characteristic examples.
 108 Chapter Five
The Polyphony of Heisenberg’s Uncertainty Paper 109
attributesto the limits of ordinary discourse, so Bohr resolved the con-
senberg did not wait to achieve the desired clarity and understanding.
tradictions of quantum theory by analyzing the limits of ordinary Jan-
Driven by ambition, relieved byhis letter to Pauli, and encouraged by
guage (MacKinnon 1994). Heisenberg moved freely between these two
Pauli’s positive response, Heisenberg sent the text of the uncertainty
options, choosing his approach according to the type of audience he
paper to print. This textis, in its main points, identical with theletter.
addressed and the type of problem he encountered.
No wonderthe uncertainty paperis shot through with doubts and con-
In his letters preceding the formulation ofthe uncertainty principle,
Heisenberg expressed his desire “to get clever” more than once. Yet tradictions. 7
Heisenberg had nodefinite notion’of an interpretation, nor did he
the sense of this expression changes:originally Heisenberg’s dominant
need one. All he intended,as he wrote in the abstractto the uncertainty
voice wasthat of denial—denial that a quantum ontology waspossible
paper, was.”’to show how microscopic processes.can be understood”
(the only enlightenmentone could hope for was to understandthe tran-
through quantum mechanics (1927b, 62). Heisenberg’s stated goal was
sition from the macro- to the microdomain); after Schrédinger’s chal-
to “illustrate” quantum theoryby discussing thought experiments. Yet
lenge, Heisenberg, more ambitiously, began the search for deeper un-
in orderto disseminate the newtheory effectively and to makeit palat-
derstanding. In a broad sense, the Gottingen-Copenhagen physicists
able, Heisenberg had to offer not simply a satisfying “illustration” but
sought to understand the “naturalness” of agreement between theory
one that was intuitive (anschauliche). How does one claim intuitiveness
and experiment. For Born, a Leibnitzian preestablished harmony ex-
for.such an abstract, nonintuitive formalism? At the very end ofhis let-
isted between the world of mathematical symbols andthat of facts ob-
ter to Pauli, Heisenberg madea hesitant suggestion: perhaps onecan
servedin the laboratory. Born, as I have argued, used the notion of pre-
say that one understands physical laws in an intuitive way when one
_ established harmony to justify his own statistical interpretation. This
can immediately say in each experimental case what should occur—in
notion was retained in Bohr’s Como lecture, which was woven around
this way the sense of.understanding is advanced (Heiseriberg to Pauli,
the “harmony” between definition and observation (see: chapter 6). 27 February 1927, PC). After Pauli provided his support, this hesitant
Physicists outside the mainstream, such as Campbell, also suggested
suggestion was transformedinto the confident redefinition ofintuitive-
that to substantiate a radically new theory, one should discuss experi-
ness with which Heisenberg openedthe uncertainty paper: “We believe
ments (thought experiments and actual experiments) and demonstrate
that we intuitively understand physical theory when wecan think
howthe theory works (how the theoretical concepts are used in experi- © -
qualitatively about individual experimental consequences and at the
mental situations; chapter4).
same time we knowthatthat application of the theory never contains
Heisenberg’sletter to Pauli is built around “making sense”out of the
internal contradictions’ (1927b, 62). The justification for this newdefi-
uncertainty formula in numerous thought experiments (Heisenberg to
nition, as is clear from the Heisenberg-Pauli correspondence, is that
Pauli, 27 February 1927, PC).” Heisenberg’s discussion of these experi-
Heisenberg’s qualitative discussion “eased his conscience” and simul-
ments in the letter convinced him that he was able to offer a satisfying
taneously provided an effective way to counter Schrédinger’s criti-
interpretation of quantum physics. Despite his continuing doubts, the
cisms. Still, as we will shortly see, Heisenberg was not completely sat-
discussion had “eased” his conscience (mein Gewissen erleichtert). In-
isfied with this definition, offering, in addition, a visualizable
deed, Heisenberg had openedtheletter to Pauli by expressing a hope
interpretation of the uncertainty formula in termsof the directions of
that by the very act of writing he would clarify the interpretive issues
tensors (matrices) in Hilbert space, thus endowing these abstract enti-
to himself (dadurch hoff’ ich auch selbst driiber klar zu werden). Yet the feel-.
ties with ontological meaning. These two conflicting notions of An-
ing ofrelief and insight that Heisenberg achieved waslimited: he con-
schaulichkeit run throughthe uncertainty paper.
cludedthe letter by saying thathis deliberations might be merely vague
speculations, that they probably contained nothing new, and
We can conclude that Heisenberg did not start with a clear concep-
that they tion of whata satisfying interpretation should be; rather, the decision
were unclear on many points. In order to achieve clarity, Heisenberg
about the adequacyof his notionof interpretation was conditioned by
revealed, he would need further communication with Pauli.®
Yet Hei- social pressure (for Anschaulichkeit) and Heisenberg’s own feeling of
7. The existingliterature on the history of quantum mechani
cs usually only mentions relief (that his “conscience” waseasier). Pauli, whose “relentless criti-
oneof these thought experiments—the y-ray experiment. cism” Heisenberg requested (Heisenberg to Pauli, 27 February 1927,
8. “To make anything clearer, I must write to you aboutit” (Heisenb
27 February 1927, PC).
erg to Pauli, PC), had a more demanding conception of an interpretation. If Hei-
:
senberg put the uncertainty formula in the center of his interpretive
 110 Chapter Five
The Polyphonyof Heisenberg’s Uncertainty Paper 111
enterprise, Pauli expected that the uncertainty formul
a should illumi- We can only speculate about what caused Heisenberg to
nate directly the nature of the basic quantitative laws of quant turn to this
um me- issue at the last moment. Did he fully realize the consequencesofhis
chanics. Heisenberg quite confidently disagreed: Pauli
was asking too dialogue with Jordan, along thelines of the analysis in chapter 4? Did
much (Heisenberg to Pauli, 9 March 1927, PC). One
encounters here, he suddenly remember the challenge of his teache
according to Heisenberg, the same theoretical freedom r Sommerfeld (1927),
asin all other

ton
physical theories—quantum theoryis no exception. In relativ who awaited a new Kant who would overthrow classical
causality, as
ity theory, Einstein had disposed of Newtonian space-time? Did Heisen
Heisenberg argued, one cannot substantiate the principle berg real-
of the con- ize that an essentially statistical theory, which is mute on questions of
stant velocity of light—it is merely the simplest assump
tion if one ac- individual phenomena,is always open to a charge of incompletene
_ cepts Einstein’s definition of simultaneity. Similarly, if one ss,
knowsthat Pp ‘thus weakeninganyclaim for the finality of the Gottingen-Copenhagen
and q are not regular numbersbut that therelation Pigi
~ (h/ 277i) holds, achievement(an achievementto which Heisenberg had contributed so
then the assumptionthat'p and q are matricesis the simplest.?
‘Heisenberg, despite his confident tone, was aware decisively)? Did Heisenberg read, hear, or see at the last moment
some-
that such an in- thing of the acausalspirit prevalent in his cultural milieu? We probably
terpretation isnotself-evident: "Naturally, I clearly see
that this for- will never know the answer. Yet realizing how contingent the introduc-
mulation can appear unsatisfactory” (Heisenberg to Pauli, 9
March tion of the acausality issue into the uncertainty paper was, and how
1927, PC).° Still, he retained his arguments, he informed Pauli,
as a con- closely tied to the original content of the paperit later became, is very
clusion of the uncertainty paper. This conclusion, Heisenberg admitt
ed, instructive. . : ‘
is too dubious, too uncertain—he might yet write “ten times”
(!) some- Whatis the ultimate reason for Heisenberg’s “final” overthrow of
thing completely different, or he might delete these arguments from
the causality? Themere fact of employing a statistical formalism for the
final version altogether. As it happens Heisenberg did not delete these
derivation of the uncertainty principle does not, of course, signify any
arguments. So much for the story of his strong “belief” and “commi
t- predilection for “essential” acausality —this is the reason Born,
ment” to the arguments in the uncertainty paper. Jordan,
and Pauli all considered the problem of causality open before Heisen-
berg’s uncertainty paper. Double-voicedness characterizes Heisenberg’s
The Contingency of Acausality discussion ofthe causality issue in the middle of the paper—it is be-
cause Heisenberg added the verdict on causality at the endof the paper
Heisenberg’s uncertainty paper has won a permanentplace in
the an- that the contrast betweenthe voices is so strong. Is quantum mechanics
nals of science andin the history of philosophybecause ofits
radical “characteristically”’a statistical theory, as Born andJordan claim? Oris
message ofthe “final” overthrow of the lawof causality. In Wheele
r and statistics only “brought in by experiments,” as Dirac maintained? No
Zurek’s (1983) anthology it is reprinted underthe heading “Princi
ple of preference is spelled outinitially.” In a letter to Pauli, Heisenberg had
Indeterminism.”
mentioned noncommittally that Dirac’s stand seemed “deeper” to him
Yet Heisenberg must have introduced his “final” overth
row of cau- (Heisenberg to Pauli, 27 February 1927, PC). At the end of the uncer-
sality at the very last stage of writing the paper, becaus
e the causality tainty paper Heisenberg confidently ruled out the “essentially statisti-
issue is mentioned neither in the outline of the paper contai
ned in the cal” nature of quantum theory (citing the Bothe-Geiger experiments
27 February letter to Pauli, nor in two subsequentletters, which
further against it), when “onlystatistical conclusions can be drawn from pre-
discussedissuesin the paper. Noris the acausality problem
mentioned cise initial data” (1927b, 83). Heisenberg’s resolution was to bring into
in the abstract.
full view the impossibility of knowing the present, andtherefore of pre-
9. Heisenberg repeats this argument in the uncerta dicting the future, thus disposing of any hopesof determinism.
inty paper without mentioning
Pauli: “Of course, we would alsolike to be able to deri ve,
if possible, the quantitative laws
How can one argue that the present cannot be known exactly? One
of quantum mechanics directly from the physical foundat
ion—thatis, essentially, from can point out, of course, that it follows from the uncertainty formula,
relation (1) [p.q,
~ h]. Webelieve, rather, that for the time being
the quantitative laws can
be derived outof the physical foundations onlyby the
use of maximum simplicity. If, for 11. “In the fact that in quantum theoryonly the probability distribution of the
example, the X-coordinate of the electron is no longer a position
number .. . , then the simplest of the electrons can be given for a definite state, such as 1s, one can
assumption conceivable {that does not contradict recognize, with Born
relation (1)] is that this X-coordinate is andJordan,a characteristicallystatistical feature of quantum theory as contrastedt
a diagonal term of a matrix” (1927b, 82). o clas-
10. “Natiirlich, ich seh’ gut, dass diese Formulieru sical theory. However, one can say, if one will, with Dirac, that the
ng unbefriedigend scheinen kann.” statistics are brought
in by our experiments” (Heisenberg 1927b, 66).
 112. Chapter Five
The Polyphony of Heisenberg’s Uncertainty Paper 113

whichis a mathematical consequence of quantum mechanics. The per-

suasivenessof the overthrow of causality thus dependsonthestatus of the uncertainty formula “creates room’for the validity of the quantum
mechanical commutation relation: if position is determined precisely,
the “finality” of quantum theory. In 1927, when this highly abstract and
momentum is undetermined. This qualitative “making sense”is, of
unconventional theory had yet to prove its mettle, such an argument
course, different from straightforward mathematical derivation. Heisen-
would convince only believers.
bergderived the uncertainty relation while supplementing the original
Heisenberg therefore needed arguments that would show “howthe matrix formalism with thestatistical assumptions of the Dirac-Jordan
world is the way quantum mechanicssaysit is” withoutdirect appeal
to the formalism. This is the reason Heisenberg preferred Dirac’s stand, transformation theory—no wonderhegetsa statistical result! This was
one of numerousstrategies used, intentionally or not, to present the
seeking to understand how “statistics is introduced by experiments.”
Gottingen-Copenhagenperspective as unassailable.
Heisenberg’ssearch resultedin the introduction of the (misleading)dis-
turbance imagery—with the help of disturbance imagery Heisenberg
Anschaulichkeit and the Status of Classical Concepts
arguedthat there are no continuousorbits in an atom, just as the quan-
tum mechanical formalism implies. One could deducethis fact “with-
As Ihave mentioned, Heisenberg was torn between his desire to render
out knowledgeof recent theories, simply from experimental possibili-
matrix mechanicsintuitively appealingto a wide audience and his in-
ties” (1927b, 65). Consequently, the impossibility of a causal space-time
fatuation with Dirac’s abstract version of quantum theory. This conflict
description seemed to follow directly and inevitably from the condi-
. was not resolved: it permeates the uncertainty paper. Heisenberg vac-
tions of experience, as Bohr later claimed.
illated between his attempts to “intuit,” or “understand,” quantum
Two voices characterize Heisenberg’s discussion of causality, An-
theory by discussing simple thought experiments and his confident de-
schaulichkeit, and otherissuesin the uncertainty paper. These conflicting
duction of philosophical consequences from the mathematical formal-
voices perseverein later writings by the Géttingen-Copenhagen physi- ism of the theory. -
cists, often changing their pitch but never completely fading away. One
How does one endowthe behaviorof microscopic objects with intui-
voiceis that of “inevitability”: through simple analyses of experiments
tiveness, when onerejects their causal space-time description? What
one can see that the state of affairs claimed by quantum theory is
happensto objects located in space-time? Eager to resurrect Anschau-
unavoidable. The other voice is that of “consistency”: analyses ofex- -
lichkeit, Heisenberg found a brilliant solution: the microobject, he de-
periments demonstrate the consistency of quantum theory because the
clared, is nothing butthetotality ofall its interactions.’? In this way, one
results of such analyses do not contradict the quantum mechanical for-
can define the attributes of atomic objects by their interactions, or mea-
malism. While inevitability arguments do not withstand close scru-.
surements— this is in fact an operational stand that precedes Bohr’s re-
tiny (see chapter 9 and Beller 1993), they served as a powerful tool for
lational definition of concepts. In this way Heisenberg defined the po-
legitimating and disseminating the Copenhagen orthodoxy. Inevitabil-
sition of an electron through measurements by a y-ray microscope and
ity arguments are often mixed with those of consistency. Heisenberg’s
redefined the path of a particle as a sequence of measurements. Simi-
concluding words in the uncertainty paper are uttered in the moreso-
larly, Heisenberg defined the “size” of an electron as the difference in
ber, consistency voice. Inevitable acausality follows from the correct-
the electron’s position when twoveryfast particles hit it one after the
ness of quantum mechanics (which covers the measurement process as
other within a very short time interval At (1927b, 65).
well): “Becauseall experiments are subject to the laws of quantum me-.
Heisenberg was considerably moreproficient than Bohr in working
chanics and therefore to equation (1) [p,q, ~ h], it follows that quantum
with the mathematical tools. of the quantum mechanical formalism.
mechanics establishesthe final failure of causality” (1927b, 83).
Heisenberg had noreasontherefore to give up his attempt to explore
Heisenberg also employed anotherstrategy in establishingthe inevi-
the ontological significance of the quantum formalism. Consequently,
tability of acausality. He claimed that the indeterminacy relation is
“the two voices permeate Heisenberg’s writings—the positivist, identifying
real basis for the occurrenceofstatistical relations in quantum mechan-
ics” (1927b, 62). He promised to demonstrate that the uncertainty for- 12. “In order to be able to follow the quantum-mechanical behavior of any object one
mula “is a straightforward mathematical consequence of the rule has to knowthe massof this object andits interactions with any fields and other ob-
pq —
qp = —th” (1927b, 65). Yet Heisenberg’s promise was not kept. jects: ... About the ‘Gestalt’ (construction) of the object any further assumptionis unnec-
What essary; one most usefully employs the word ‘Gestalt’ to designatethetotality of these
Heisenberg wasable to offer was merely a qualitative discussion of how
interactions” (Heisenberg 1927b,64).
 114 Chapter Five
The Polyphony of Heisenberg’s Uncertainty Paper 115
the meaning of a concept with the procedureforits verification
(in real stract character of electromagnetic theory as an argume
or imagined physical interactions), and the realisi, deducing nt against the
the genu- temptation to look for an intuitive interpretation.
ine features of the quantum world from characteristics of the mathe- The fact that such
matical formalism. The invariant features of this formalism, features of the quantum formalism as, say,
according the terms of a position
to Heisenberg, explain why the uncertainty relations must hold. matyix can be connected through bridge principles with
Inorder a position
to obtain physical results from Dirac’s formulation of quantum theory, measurement is not an argument for Anschau
lichkeit as Heisenberg’s
one hasto associate ordinary numbers with the q-numb contemporaries understoodit. Even less is it an argumentfor the “in-
ers(matrices, or dispensability of classical concepts,” Bohr’s notorious
_ tensors, in multidimensional space). A definite experimental setup doctrine. Heisen-
pre- berg, let us note, did not argue in the uncertainty
_ scribes a definite directionin’ this multidimensional space—the paper for the “indis-
ques- pensability of classical concepts.” On the contrary, he claimed
tion ofthe “value” of the matrix in this direction has a well-defined that in
meaning only when the given direction coincides with the direction the case ofrelativity theory, “the possibility of employing usualsp
of ace-
a principal axisof the matrix. When these twodirectionsdiffer slightly time conceptsat cosmological distances can bejustified neither by
logic
from one another, one can still talk about the “value” within a certain nor by observation” (1927b, 62). Similarly, in the microdomain
“the ap-
plicability of classical kinematics and mechanical concepts
statistical error. A definite experiment therefore cannot give exact infor- ... can be
justified neitherfrom our lawsof thought nor from experiment” (1927b,
mation about all quantum theoretical quantities —ratherit divides the
82)."* Yet becauseHeisenberg resurrected in the uncertainty paper the
“physical quantities into “known” and “unknown,”asis exemplified in
kinematic interpretation of matrix theory (as opposed to the
the uncertainty formula.”In this way Heisenberg “makes sense” out of earlier
the abstract quantum formalism. electromagnetic interpretation) and because he analyzed thought ex-
periments in termsof classical space-time concepts, the impression
Heisenberg’s fearless exploration of quantum reality was intended can
arise that he was aligning himself with the Bohrian doctrine of their
not merelyto gain deeper physical insight into the workings of the
quantum world. Heisenberg seemed neverto be free from a desire to indispensability. a
The tension between Anschaulichkeit and Unanschaulichkeit is mir-
argue—again and again—the superiority of matrix mechanics to wave
mechanics. He offered in this respect a peculiar mixture of positivist rored in the discussion of the applicability of space-time concepts with
(using disturbance imagery) andrealist (using ontological discuss
which Heisenberg opens the uncertainty paper. We clearly discern
ion) there many conflicting voices, including Heisenberg’s own pastvoi
argumentation. Thus Heisenberg claimed: “The X-coordinate of the ce.
electron is no longer a ‘number,’ as can be concluded experimentally
Thus Heisenberg repeated his conclusions from the earlier Naturwis-
,” senschaften paper (1926b) that no contradiction-free interpretation
and in this case, the “simplest assumption conceivable . . . is that the X- of
quantum mechanicsin termsof the usual kinematic concepts is possibl
coordinate is a diagonal term of a matrix” (1927b, 82). Heisenberg did e
not show whythis is the “simplest assumption conceivable’”—he im- because of the internal discrepancies that “show themselves in argu-
ments aboutcontinuity versus discontinuity and particle versus wave”
mediately proceeded to defend this new and abstract kind of reality,
_ countering a potential critical interlocutor, perhaps Schrédinger: “The (1927b, 62). This argumentis consistent with the original philosophy
prediction that, say, the velocity in the X-directionis ‘in reality’ not a of matrix theory, which arose “exactly out of the attempts to break with
all ordinary kinematic concepts” (1927b, 62). Yet, rather than deduci
number but the diagonal term of the matrix, is perhaps no more ab- ng
the need to renounce space-time geometry, as he originally did, Heisen-
stract and no more unvisualizable than the statementthat the electric
berg now announcedthat revision of space-time geometry
field strengths are ‘in reality’ the time part of an antisymmetric tensor at small
located in space-time. The phrase‘in reality’ here is as much and as little distances is “unnecessary,” thus implicitly disputing with Campbell,
justified as it is in any mathematical description of natural process Sentfleben, and his own pastvoice. The reason for this conclusion, ac-
es” cording to Heisenberg,is that: “We can make the quantum-mechanical
(1927b, 82). :
Heisenberg’s “intuitive” interpretation turns out to be notintuitive
at 14. Later, Heisenberg would attimesalign himself with Bohr’s
doctrine, introducing
all. Jordan, just a few monthsearlier, contra Heisenberg, used
the ab- manyinconsistencies into his arguments. See chapters
8 and 9.
15. No wonder Heisenberg wasnoteagerto accept Bohr’'s insistenc
e on the indispen-
13. For every quantum theoretical quantity one can find a coordina sability of both waves and particles. The appendedpostscript, where
te system in which Heisenberg sup-
the quantity has an exact value. ported Bohr’s position, standsin striking contrastto these openinglinesof
é the uncertainty
paper.
CRRATT
116 Chapter Five

laws approximate theclassical onesarbitrarily closely by choosing suf- CHAPTER 6

ficiently great masses, even whenarbitrarily small distances and times
come into question” (1927b, 62). Heisenberg did not substantiate this
ES
His

claim, which again contrasts with his reasoningjust a yearearlier.

Still, the revision of kinematic and mechanical concepts is necessary,
The DialogicalBirth of
Heisenberg claimed, because such a need follows from the basic rela-
tion pq — gp = —ih (1927b, 63). These two voices are not easy to har-
Bohr’s Complementarity
monize: one calling for the resurrection of regular space-time geometry
(so that thought experiments canbe discussed), the other demanding
His [Bohr] turn ofmind was essentiallydialectical, rather than
the revision of kinematic concepts, such as position and momentum (as
reflective. . .. He needed the stimulus of someform of dialogue to start
dictated by the mathematical formulas). In Heisenberg’s paper, they his thinking. —
simply coexist and constitute yet another source of inconsistencyin his
writings (on this score the original philosophy of the matrix theorists Leon Rosenfeld 1967, 117
was more consistent).
In the uncertainty paper Heisenberg argued that in a discontinuous It ts, of course, possible to simplify the medium in which a scientist works by
world “contradiction between the concepts of ‘position’ and‘velocity? . simplifying its main actors.. :
is quite plausible” (1927b, 63). He substantiated this claim by arguing
‘Paul Feyerabend 1975, 19
that in a continuum theory the velocity of a particle is given by tangents.
to the position curve at every instant. In a theory based on disconti-
nuity, we haveinstead ofthis curvea series of discrete points. In sucha
case, Heisenberg claimed,“it is clearly meaningless to speak about one
velocity at one position (1) because one velocity can only be defined by
ee two positions and (2), conversely, because one point is associated with
Introduction
i two velocities” (1927b, 63). Se
This intuitively appealing argumentis problematic—what meaning A dialogical analysis of a scientific paper differs from the usual concep-
can oneattach to a discontinuous world in which points are distributed
tual analysis. The dialogical approach is opposed to static structures
with finite separation, unless one presupposes a continuous space-time - and fixed meanings—itis inherently contextual and historicist.
container underlying the discontinuous one? Here Heisenberg’s rea-
Such an analysis, as we sawin the previous chapter, can change our
i soning contrasts with the inspiration of the original philosophy of understanding of a scientific text: we uncover diversity of meaning,
bE complexity of argumentation, unresolved tensions. When weperceive
matrix mechanics, which arose precisely from attempts to eliminate
that the text is populated with invisible interlocutors, we realize what
the superfluous continuoussubstratum altogether.’* Ingenious Heisen-
berg attempts the impossible:to show,by simple.intuitive” arguments the central issues of the paperare. This realization modifies our under-
standing of the content of the paper. Dialogical analysis is a potent
yoreTere tsi tyr’ ty

and without appeal to the quantum mechanical formalism, why quan-

%

tum mechanics is “inevitable” or, at the very least, “quite plausible” tool for deciphering opaque and obscuretexts. Such is the case, I argue
(1927b, 63). in this chapter, with respect to the analysis of Bohr’s Como lecture
(1927c)—Bohr’s initial formulation of his-complementarity principle.
16. Matrix theorists, claiming to have established a “truly discontinuous theory,” ob- The usual reading of the Comolectureidentifies the announcement
jected to the quantization procedure that chose the quantized mechanical orbits of an of wave-particle duality as the main message of Bohr’s initial presenta-
atom from all possible mechanical ones; a satisfactory discontinuoustheory,
so the origi- tion of complementarity. My dialogical reading challenges the accepted
nal argument went, should not contain the unrealizable motionsatall. reading: Bohr continuesto reject pointlike light quanta, and his discus-
sion is heavily asymmetrical in favor of waves. The usual reading of
Pree SoDPTR ONM Tp at 8

Bohr’s paper assumes the similarity of Bohr’s and Heisenberg’s posi-

tions, while a dialogical analysis reveals an incompatibility between
 118 Chapter Six
The Dialogical Birth of Bohr’s Complementarity 119
their positions at the time. The usual reading also assumesthe
central- Bohr’s first presentation of the complementarity principle took
ity of measurements and of operational definitions place
of concepts, while at the International Congressof Physics, held in the Italian
the dialogical reading of the Como lecture uncovers Bohr’s city of Como.
more mod- In the history ofscientific thoughtit is hard to find another contri
est emphasis on the “harmony” between the wavetheoret bution
ical defini- about which opinions continuedto differ sharply more
tion of concepts and thepossibilities of observation. than half a cen-
Significantly, the tury after its appearance. Somephysicists, such as Leon Rosenf
dialogical reading discloses that the central message of eld, con-
Bohr’s paper sidered complementarity the most profoundintellectual insight of
was notthe resolution of wave-particle duality by the
the complementarity twentieth century, the pinnacle of the physical understandingof
principle but rather anextensive defense of his concep nature,
tof stationary no less inevitable than “the emergence of man himself as a produc
Ee ee

states and discontinuous energy changes, or “quantum jumps t

at}

.” of organic evolution” (1961, 384). Others criticized Bohr’s complemen-

_ As with Heisenberg’s uncertainty paper, plurality
of meaning per- tarity as an obscure “double-think” that impeded clear thinking and
meates Bohr’sinitial elaboration of complementarity
. An analysis of scientific progress, or as a crutch that helped initially but was even-
Bohr’s discussion of the complementarity between space-
time and cau- tually no longer needed (Landé 1967; interview with
sality reveals different, even incompatible uses ofthis Alfred Landé,
concept. Schol- AHOP).
ars have attempted to find an unequivocal connection between wave-
Attitudes toward Bohr’s complementarity principle evenin the camp
particle complementarity and space-time—causality compl
ementarity of “believers” were more ambivalent than the published sources dis-
(Murdoch 1987). Some have arguedthat causal descriptions associ
ated close. As Dirac expressedit: “I never liked complementarity. ... It does
with particles and space-time with wave propagation, while others
con- not give us any new formula.. . . I believe the last word wasnot said
versely connect space-time with particles and causality with waves.
In yet about waves andparticles” (interview with Dirac, AHQP). Simi-
fact, both of these contradictory readings are present in the text.1
They larly, Heisenberg disclosed, “I know that, besides Landé, many other
occurin different dialogues and address differentissues. Compl
emen- physicists had been upsetby this situation, and they felt it was a dual-_
tarity between space-time and causality is an imprecise
umbrella con- istic description of nature” (interview with Heisenberg, AHQP).
ceptthat allows Bohrto copelocally with interpretive issues while
en- Despite the strongreactions it provoked, what exactly Bohr’s comple-
trenchinghis initial conception of stationary states and discon
tinuous mentarity means continues to be an enigma. Jammer, commenting
on
energyjumps.
the difficulty of comprehending Bohr’s complementarity principle, de-
A dialogical reading of Bohr’s Comolecture also brings
back to life scribed Carl Friedrich von Weizsacker’s extensive effort to elucidate
“lesser”scientists. We will see that the work of Campbell trigge
red the original meaning of complementarity. However, when Weizsacker
Bohr’s deliberations oninterpretive issues as it did Heisenberg’s
(chap- asked Bohr “whetherhis interpretation accurately presents what Bohr
ter 4). Campbell's namefigures in a draft of Bohr’s paper (1927b
, 69), as had in mind, Bohr gave him a definitely negative answer” (Jammer
well as in the published text (1927c, 131).
1974, 90). Recent students of Bohr’s thought continue to face the “for-
TET wr i te sf eer

A dialogical reading of Bohr’s Comolecture also provides a new

per- midable difficulty,” even incomprehensibility, of his writings dealing
spective on the famous clash between Heisenberg and Bohr
2h

over the with complementarity. ThusFolse claimed, “Even with almost one full
uncertainty paper. A dialogical analysis reevaluates Bohr’s
and Heisen- year’s work onthis paper, innumerable rewrites, two public deliveries,
berg’sintellectual positionsso that the confrontation between
them be- and three different sources of publication, the essay contains many ob-
comes conceptually meaningful, no longer merely the
result of “mis- scurities and never makes very clear what this new ‘viewpoint’ is sup-
understandings” and “confusion.” Such a presentation. allows
us to posed to be”(1985, 108).2 Don Howard recently proclaimed that “now
merge the “conceptual” and “anecdotal” aspects of the
history of quan- there are signs of growing despair ... about our ever being able to make
tum physics (Beller 1996b).
good sense outof his [Bohr’s] philosophical view” (1994, 201).
1. While discussing the Compton effect, Bohr associated A dialogical perspective allowsus to decipher Bohr’s philosophyin
conservation laws—causal-
ity —with particles and space-time descriptions with wave
propagation. Yet while dis-
general and the meaning of Bohr’s original presentation of the comple-
cussing the complementarity between Space-time and
causality in the case of stationary 2. Similarly pessimistic is the conclusion of Gibbins: “Niels Bohr wentto great lengths
states of an atom, Bohrassociated causality with waves(a
stationary state is described by to refine and thento clarify his thoughts on quantum mechanics, but, sad to
a proper vibration and hasa definitive unchangingvalue say, his
of energy) and space-time de- writings do not appearto have benefited from these efforts. It is very difficult
scriptions within the atom with particles. See the analysis below. to say...
howhis ideas hang together” (1987, 53).
ree
 120 Chapter Six
The Dialogical Birth of Bohr’s Complementarity 121
mentarity principle in particular. Without identifying the interlocutors
packet is particularly suitable for demonstrating the harmony between
of each sentence of the Comolecture,it is impossible to understand the the possibilities of wave theoretical definition and those of observation
meaning of these sentences and the connections among them. Yet when
(Bohr 1927d). The agreement claimed between the possibilities of defi-
werealize thatthe textis filled with implicit arguments with the leading
nition and observation wasinitially much weaker than the operational
physicists of the time—Einstein, Heisenberg, Schrédinger, Compton,
assertion that measurementis the primary and indispensable part of the
Born, Dirac, Pauli, and the lesser known Campbell—the fog lifts and
interpretation of concepts—thestandlater taken by Bohr. Mere “har-
Bohr’s presentation becomesclear. mony” between observation and définition of concepts is, of
A dialogical reading reveals that the central message of Bohr’s Como course, a
necessary part of anyinterpretive attempt?
lecture was the announcement of a compatibility, indeed, a “happy
~Myclaim that Bohr‘s original announcementof complementarity was
marriage,” between Schrédinger’s successful continuous wave me-
not a symmetrical solution of the wave-particle dilemma, and thatin all
chanics and the quantum postulate, especially Bohr’s original notion
physically interesting cases—those of interactions—one must neces-
of discrete stationary states. In Bohr's presentation, the de Broglie—
sarily deal with wave packets(light waves or matter waves), will be
Schrédinger wave packet was sufficient to resolve the paradoxes of
substantiated by discussing Bohr’s dialogues with Einstein, Compton,
atomic structure, of the interaction of radiation with matter, and
of the and Campbell. It is generally assumedthatthe roots of Bohr’s comple-
interaction of matter with matter—paradoxes that had confounded
mentarity lie in the experimental refutation of the Bohr-Kramers-Slater
Bohr since hisbrilliant and troublesome debut into atomic theory in
theory. This theory used the wavetheoretical framework of light exclu-
LT TRGSE

1913. The existing accounts of complementarity center almost excli-

sively rather than Einstein’s theory of light quanta. After the Bothe-
sively on the “wave-particle duality” of a free particle, while the signifi- Geiger experiments, many authors assume, Bohr had no choice but to
_ cant physical problems—thosethat invoke physicalinteractions (bound
assimilate light quanta, which hehad rejected vigorously beforehand.
electrons in atoms, collisions)—are neglected. Yet it was the paradoxi-
The foundations of Bohr’s complementary framework, they hold, lay in
cal, nonclassical aspects of these problemsthat were central to the ef-
his acceptanceoflight quanta in 1925, beforeall the subsequent break-
forts of physicists at the time. Te , throughs—Heisenberg’s matrix mechanics and Schrédinger’s wave me-
The Como lecture announced resolution of these long-standing
chanics—occurred (Jammer 1966; MacKinnon 1982).
problems by harnessing the wavetheoretical framework to the
quan- Targue for a substantially different perspective. Bohr did not adopt
tum postulate. In fact, each of the dialogues centers on Bohr’s solution
the idea of pointlike light quanta (as used in the explanation of the
of oneof these problems— atomicstructure, the interaction of radiatio
n — Compton effect), even after the Bothe-Geiger experiments. Bohr’s com-
and matter, and collisions. In eachcase, Bohr implied that the abstract
plementarity principle implied further rejection, not acceptance, of the
idea of a point-massparticle is inadequate and must be replaced by a
idea of pointlike material particles. It is incomprehensible that Bohr
wave theoretical superposition of light waves or matter waves. The fer-
' would have developedhis interpretation of quantum physics without
tility of this idea for Bohr’s skillful weaving of the argumentation in
responding in a significant wayto the pivotaltheoretical developments
the Comolecture clearly demonstrates why, from the very beginning,
that occurred during the years 1925-27. I will discuss the nature of
he ascribed physical significance to Schrédinger’s theory, unlike his
Bohr’s response in the Comolecture to Heisenberg’s matrix mechanics
younger colleagues Heisenberg and Jordan. Bohr’s implicit
argument and to Schrédinger’s wave mechanics. In particular, I will describe
for the superiority of wave concepts over particle concepts. does
not Bohr’s understanding of the space-time problem in the interior of the
mean that waves represented for him literal picture of reality. Rather
atom, and his defense and elaboration of his idea of a stationary state.
the fruitfulness of the wave modelis in the English tradition:
the wave This description further substantiates myclaim thatit is around Bohr’s
model is both heuristically useful.and theoretically adequat
e. Bohr’s concepts of stationary states and quantum jumps—which Schrédinger
Como lecture constituted a striking contrast to Heisenbe
rg’s efforts to aimed to eliminate— rather than around the wave-particle dilemma or
develop an exclusively corpuscular ontology.
The fruitfulness of the concept of wave packetsis especial
indeterminism, that the newphilosophy of physics was erected. Bohr’s
ly evident full-fledged defense of his ownideaofstationary states is elaborated in
in the case of interactions. Because we neverobserveeither
an isolated
particle or a monochromatic wave(both are “abstractions”),
but only 3. This was argued by both Born and Heisenberg in 1927for their respectivestatistical
cases of superposition of light or matter waves, the
idea of a wave and indeterminacyinterpretations.
 122 ChapterSix
The DialogicalBirth of Bohr’s Complementarity
123
paragraph of the Comolecture—a part either neglected
or considered In a manuscript written just a few days befor
the most obscure (the opinion of the editors of Bohr’s e the Como
lecture (Bohr
collected works; 1927a), Bohrintroduced de Broglie wave packe
BCW, 6:30). ts to representboth light
Bohr’s defense and elaboration of the idea that an atomic system quanta and electrons. Bohr described de Brogli
is e’s ideas “of ascribing a
adequately represented by a sequence of stationary states frequency to anvag encycarrying energy” and consequently
that are, in “a phase-
turn, adequately described by Schrédinger’s wave function wave to a material particle.” Accordingto Bohr,
reveals a the matter wave theory,
deep conceptual gap between Bohr’s wave theoretical and Heisen or “the representing of a particle by mean
berg’s s of a wave-packet, represents
particle-kinematic interpretations of atomic systems—a a direct generalization ofthe light quantum theor
gap that was y.” Such a point of
circumvented rather than resolved by subsequent develo view strongly suggests not only that the point
pments. This like corpuscular light
incompatibility between the positions of Bohr and Heisenbergis quantum is an abstraction butthat a singular harm
one of onic light waveis
the historical roots of the inconsistencies that plague the Copen similarly merely an abstraction thatis never actua
hagen lized byitself in any
interpretation of physics. My discussion of this gap also physical situation: “As emphasized by de Broglie the
provides an abstract character
insight into Einstein’s and Schrédinger’s early dissatisfaction with of the phase-waveis indicated by the fact that its
the velocity of propa-
gation . ..isalways larger than the velocity of light c.” All this sugge
Copenhageninterpretation. I argue that their initial criticism focused sts
on the inconsistency of amalgamating the incompatible positions held also that “a light wave may be considered as an abstractio
n, and that
reality can only be ascribed to a wave group” (Bohr 1927a,
by Bohr and Heisenberg.Part: of the incomprehensibility of the Como 78, myitalics).
lecture derives from Bohr's attemptto conceal this gap by uniting forces Recognizing the physical significance that Bohr attached
to a wave
packet allows us to see the following often-quoted sentencein
against the opposition. a differ-
ent light: “Radiation in free space as well as isolated materi
al particles
are abstractions. . . their properties on quantum theory being
Dialogue with Schrodinger: The Structure of Atoms observ-
able and definable only through their interaction with other
systems” |
Mydiscussion of Bohr’s predisposition to de Broglie—Schrédinger (Bohr 1927c, 116).
con- This sentence does not imply, or at least did not originally
cepts is based on the text of the Como lecture (Bohr 1927c), together imply, a
with two manuscripts written before the lecture (Bohr 1927a, strong instrumentalist approach, where noreality
1927b). can be ascribed to
atomic systems in themselves—the view Bohr gradually devel
I will also quote from a manuscript (1927d) written only three weeks oped by
countering later objections. Neither doesit indicat
after the Como conference. These three manuscripts are conveniently e a symmetrical so-
available in Bohr’s Collected Works, volume 6.4 As is clear from lution of the wave-particle dilemma. Rather only the idea
these of a wave
writings, Bohrin his original struggle with the physical interpretation packet represents the “real thing” that allows the untangling
of the
of quantum theoryleaned heavily on the idea of a wave packet—a su- mysteries of the atomic world.
My interpretation is confirmed by Bohr’s analysis of
perposition of wavesof different frequencies that results in a wavefield the possibilities
of observationin principle, as well as by his analysis of actual
limited in space and time. Bohr used the imagery of a wave
packet physical
whenever he described light quanta or electrons, both in cases of interactions—theinteraction of radiation and matter and the
free Davisson-
individuals andin cases of interactions between them. The position of Germer experiments. According to Bohr, any observ
ation necessarily
the light quantum is the position of such a limited wave field, rather involves superposition. In the case of light, “as stressed
by de Broglie
than thatof a mass-point.5 Usingthe idea of wave packets, Bohrdirectly the only way of observing an elementary waveis by
interference” (Bohr
derived the uncertainty relations, a derivation thatis oppose 1927a, 78). Here the possibilities of definition of a wave
d to Hei- packet and the
senberg’s, based onthe idea of pointlike electrons and photons. possibilities of observation are in close harmony. On the
other hand,
“the only way to define the presenceof the waves is throu
gh analysis
4. Page references for Bohr (1927a, 1927b, 1927c, 1927d) of the interaction between light and matter” (Bohr 1927b,
are to BCW, vol.6. 70). It
was
5. “Onlybythe superposition of harmonic wavesof different wavelen from this phenomenon thatthe corpuscular character
gths and direc- of light was de-
tionsis it possible at a given timeto limit the extension
in space of the wave-field. duced. Here, according to Bohr, the discussion “was
we ask aboutthe position ofa light quantum, wefind that no . . . If until recently most
more than in the caseofits unsatisfactory” because “the behaviourof the so-called
energy and momentum, wecan define a position of a light
quantum at a given time,
material parti-
cles rested so entirely on corpuscularideas,”that is,
without consideration of complementary waves”
(1927b, 69-70). on the idea of local-
ized pointlike particles. All the difficulties are removed
“through the
 124 Chapter Six The Dialogical Birth of Bohr’s Complementarity 125

introduction of an essential wave feature in the descriptionof the behaviour of experimentation. The wavetheoretical modelallowed Bohrfinally to
of material particles due to the work of de Broglie and Schrédinger” decipher the limits of the application of space-time to the atomic do-
(Bohr 1927b, 70, myitalics). The Davisson-Germer experiments, accord- main—a problem that had occupied him since 1913, when he himself
ing to Bohr, directly confirm the matter wavetheory in the spirit of de had introduced the incomprehensible quantum leaps into an inexpli-
Broglie—Schrédinger: “The discovery of Davidson[sic] and Germer. . . cable space-time abyss. On the other hand, the wave model enabled
prove[s] the necessity of applying a wave-theoretical superposition Bohr to resurrect his idea of the stationary states of an atom—an idea
principle in order to accountfor the behaviourofelectrons. . .. the wave that seemed to have been abandonedin the matrix approach and to be
character of the electronsis by these experiments shownjust as clearly _ in danger of elimination by Schrédinger’s attempts at interpretation.
as is the wave characteroflight. . .. As [is] well known the experiments The problem ofstationary states and the problem ofthe application
are in complete accordance with the ideas of de Broglie” (1927a, 77; my of space-time descriptionin the atomic domain are closely connected.
italics). Though reference to these experiments is omitted in the text of The idea of stationary states (when outof all classical mechanically
the Comolecture, this discussion, written just a few days before Bohr’s . realizable motions, only certain motions were assumedto be- actual-
presentation in Como, forms a necessary background for understand- ized), and especially the idea of instantaneoustransitions between such
ing his argumentin the lecture. states, implied from the very beginning a radical departure from clas-
According to Bohr, an adequatevisualizable interpretation of physics sical space-time models. Yet howradical the departure would eventu-
is possible only with the help of the idea of a wave packet: “The possi- ally become nobodyinitially grasped, not even Bohr. Both his 1913 and
bility of identifying the velocity of a particle with a group velocity in- 1918 papers show that Bohr did notrule out the possibility that the
dicates the applicability of space-timepictures in the quantum theory” mechanism of the transitions would eventually be understood. The
(1927c, 118). : concept of the stationary statés.of an atom was beset with conceptual
The idea of a wave packet allowed Bohr, following de Broglie, to ra- inconsistencies from the start (Beller 1992b). Still, Bohr and other phys-
tionalize the “irrationality” of the basic quantum relation Er = JA = h, icists used this idea, which successfully explained the stability of atoms
where “corpuscles of light’ with energy E and momentum I had the and powerfully deciphered the structure of spectral lines (Heilbron and
characteristics of infinitely extended waves(period of vibration 7 and Kuhn 1969; Jammer 1966). As far as Bohr and his followers (but not
wavelength 4). It was this irrationality that Bohr cited in his earlier work © Schrédinger) were concerned, theidea of stationarystates and discon-
as the reason for his opposition to the idea of light quanta. The concept tinuous transitions between them wasfully and definitely corroborated
of a wave packet demonstrates that one is dealing not with by the Franck-Hertz experiments.
two “rivalizing [competing] concepts” but rather with a description of In retrospect, we can see how, step by step, the conceptual price of
two “complementary sides of phenomena (Bohr 1927b, 69). The rela-
Tey I es

the utility of this idea rose. In the Bohr-Kramers-Slater theory the idea
tion AfAE = AxAI, = AyAI, = AzAI, = h indicates the reciprocal accu- of stationary states and quantum jumps necessitated such major de-
racy with which the space-time and energy-momentum vectors of such partures as nonconservation of energy and momentum, as well as the
wave packets can be defined, suggesting the complementarity of space- introduction of the strange ‘virtual field’ that an atom in a definite
time and causality descriptions. Therelation indicates “the highest pos- Stationary state emitted. The Bohr-Kramers-Slater theory also implied
sible accuracy” in the definition of “individuals associated with wave further departures from ordinary space-time visualization, for ex-
fields.” In general, the wave packet would “in the course of time be ample, in the explanation of the Comptoneffect. Similarly, the analysis
subject to such changesthat [it would] becomeless and less suitable for of the phenomenonofcollisions, whena fast-moving particle collides
representing individuals” (Bohr 1927c, 119). Here, according to Bohr,is with an atom in a certain stationary state, demanded resignation from
the source of the “paradoxes” of quantum theory.
eyes

strict conservation of energy and momentum (Bohr 1925). When Bohr

As wewill see, Bohr did not accept the particle-kinematic interpre- learned of the results of the Bothe-Geiger experiments, confirmingstrict
nee TEN a

tation of an atom in a given stationary state (as Heisenberg, Pauli, and conservation, he did not question the formulas of the wave theory of
2

Born would haveit), nor did he entertain the idea of wave packets mov- light or the adequacy of the description of atomsin terms of stationary
ing along visualizable Keplerian orbits (as Schrédinger at some point states. Rather it was a further departurefrom classical space-time mod-
had hoped). In Bohr’s interpretation ofthe interior of the atom, the pos- els for the interior of the atom that Bohr found mandatory. Bohr was
- RW: SampoTe

sibilities of visualization were closely connected with the possibilities supported in this direction by the wavetheoretical ideas of de Broglie
!i
 126 Chapter Six
The Dialogical Birth of Bohr's Complementarity 127
and Einstein. The paper Bohr wrote on these matters was published in
tions represent stationary states, then “a fundamental renunciation of
1925, just before Heisenberg’s reinterpretation paper— the paperthat
the space-time description is unavoidable” (1927c, 129). The reason for
provided a panaceaforall theills of the old quantum theory byelimi-
this is the fact that “every space-time feature of the description of phe-
nating the classical space-time container from theinterior of the atom.
nomenais based on consideration of interference taking place inside
Heisenberg’s solution, however, was too radical. Matrix mechanics
a group of such elementary waves” (1927c, 129; 1927d, 97). Stationary
could not theoretically describe the states of atomic systems and the
states, having definite energy, are adequatelydescribed by a single ele-
evolution of phenomena—the mainreasonfor the lukewarm reception
mentary wave (because “the definition of energy and momentum is at-
it received initially (chapter 2). Schrédinger indeed perceived the ma-
tached to the idea of [a] harmonic elementary wave”; 1928, 113). The
trix mechanical formalism as eliminating Bohr’s conceptof separatesta- above sentences clearly indicate that no particle-kinematic description
tionary states. Schrodinger himself attempted a continuous wave de-
can be applied to a separate stationary state in principle, and that “a
scription that would further eliminate the concept of Stationary states
consistent application of the conceptof stationary state excludes.. . any
withdefinite energies, and discontinuous transitions between these
specification of the behaviorof the separateparticles in the atom” (1928,
states, by suggesting instead a resonance modelin termsof frequencies.
112). Herelies a crucial retrospectiveinsightinto the past failures of the
As I have argued, the Géttingen and Copenhagen physicists joined
old quantum theory, which assumed visualization of these stationary
forces in response to the perceived threat from Schrédinger—hisat- states in terms of mechanicalelectronorbits. Since no time mechanism
tempt to reduce Bohr’s conceptsto the status of Ptolemaic epicycles.I:
for the description of these states can be conceived, no predictionof the
was around the adequacy of Bohr’s concepts of stationary states and time of the transition is possible. This, in turn, discloses why, for de-
quantum jumpsthatthe crucial interpretative attempts revolved. scribing these transitions, one must be content with probabilities—an-
Bohr, who accepted the great usefulness of Schrédinger’s formalism, other crucial insight into the sourceof past struggles! “ss
could notsee initially how solutions to Schrédinger’s wave equation,in The initial difficulty of reconciling the idea of stationary states with
terms of the superposition of different proper vibrations, could be rec- the Schrédinger wave function wasthefactthat the general solutionof
onciled with the idea of separate stationary states. His heated argument Schrédinger’s equation is a superposition of proper vibrations, and at
with Schrédinger, during thelatter’s visit to Copenhagenin thefall of first it seemed “difficult to attribute a meaning to such a superposition
1926, had centered on quantum jumping. After Schrédinger’s visit, the . as long as we adhereto the quantum postulate” (Bohr 1927d, 97), ac-
direction of Bohr’s efforts became clear: to achieve compatibility be- cording to which an ‘atom is always in some definite stationarystate.
tween wave theoretical ideas and the quantum postulate. As Heisen- Schrodinger regarded this difficulty as conclusive, arguing that an atom
berg putit: “Bohr realized at once that it was here we would find the exists in a superposition of several proper vibrations,.and that the ap-
solution to those fundamental problems with which he had struggled propriate characteristic of a propervibrationis its frequency and notits
incessantlysince 1913, andin thelight of the newly won knowledge he energy. Bohr’s ingeniousresolution ofthis difficulty centered on an ex-
concentrated all his thought ona critical test of those arguments which had led ploration of the compatibility of the possibilities of definition and ob-
him to ideas such as stationary states and quantum transitions” (1967, 101, servation. A superposition of proper vibrations is an adequate descrip-
myitalics). tion in the case of interactions (observations). An atom in a definite
Bohr’s Comolecture was the culmination and resolution of these ef- stationary state with precise energy is, however, a closed system, not
forts. Bohr criticized Schrédinger’s attempts to replace “the discontinu- accessible to observation, and therefore it can be correctly described by
ous exchangeof energy . . . by simple resonance phenomena” (1927c,
a single proper vibration. As such,the ideaof a stationary state becomes
127, 1927d, 97). Schrédinger’s theory must necessarily “be interpreted an abstraction, both becauseit is represented by a single wave and be-
by an explicit use of the quantum postulate,” and “in direct connec-
cause a system “not accessible to observation .. . constitutes in a certain
tion with the torrespondenceprinciple.” Moreover, “in the conception
sense an abstraction, just as an idea of an isolated particle” (1927d, 97).
of stationarystates we are . . . concerned with a characteristic applica- Despite its being an abstraction, the conceptof a stationarystateis in-
tion of the quantum postulate” (1927c, 130). Countering Schrédinger, dispensable in the interpretation of phenomena (1927d, 97). Because
Bohr asserted: “The propervibrations of the Schrédinger wave-equation “the conception ofa stationarystate involves,strictly speaking, the ex-
have been found to furnish a representation of the stationary states clusion of all interactions” (1928, 115), the constant energy value asso-
meeting all requirements”(1927c, 126; 1927d, 97). If characteristic
vibra- ciated with such states “may be considered as an immediate expression
 128 Chapter Six
The Dialogical Birth of Bohr’s Complementarity 129

for the claim of causality contained in the theorem of conservation of Here again the harmony betweenthe possibilities of observation and
energy” (1927c, 130). Thus the complementarity of space-time and cau-
definition comes to the rescue, and Bohr demonstrates with character-
sality in the quantum domain follows. The idea of an atom in station-
istic ingenuity how evenin this case his conception of stationary states
ary state as a closed system with constant energy accords well with and transitions between them can be preserved.®
Bohr’s initial introduction ofthis idea to explain the stability of matter:
The possibility of describing stationary states by means of Schré-
“This circumstance justifies the assumption of supramechanicalsta-
dinger’s wave function, in contrast with the inadequacy of matrix me-
bility of the stationary states, according to which the atom, before as
chanics on thisscore, indicated to Bohr the profound physical signifi-
well as after an external influence, always will be found ina stationary ~ cance of wave mechanics.Thefact that such a definition of a stationary
state and which forms the basis for the use of the quantum postulate” ‘state is theoretically adequate is demonstrated, according to Bohr, by
(1927c, 130). Schrédinger’s interpretive aspirations are deficient pre-—
_ Born’s work, which provided. ““a complete description ofthe collision
cisely because they cannot deal with the stability postulate, for which phenomenaof Franck and Hertz, which maybesaid to exhibit the sta-
the assumptionof stationarystatesis essential.
bility of st[ationary] states” (1927b, 71). Bohr’s enthusiasm for Schré-
Bohr devoted all of paragraph 5 of the Como lecture to arguing for dinger’s wave mechanics was manifest in his correspondence at the
the consistency of the conceptofa stationary state, Bohr’s strategy was time. Shortly after Schrédinger’s visit to Copenhagen, Bohr wrote to
to demonstrate that complementarity between stationary states and Ralph Fowler: “Just in the wave mechanics we possess now the means
corpuscular space-time descriptions—or between causality and space- of picturing a single stationary state whichsuitsall purposes consistent
‘time, for the interior of an atom—accords fully with the possibilities’ with the postulates of the quantum theory. In fact, this is the very rea-
of observation. For example, if one is to inquire about the behavior of son for the advantage whichthe wave-mechanicsin certain respects ex-
separate particles in an atom, then onehasto neglect their mutual in- hibits when compared with the matrix method.” Two days later Bohr
teraction during the observation, thus regarding them asfree. This ne- argued the samepoint in a letter to Kronig. Bohr madethis point again
cessitates a very short time of observation, shorter than the periods of in his first written attempt at interpretation leading to the Comolecture
revolution of electrons. This, in turn, implies a big uncertainty in the (Bohr to Fowler, 26 October 1926, AHQP; Bohr to Kronig, 26 October
energy transferred during the observation, and thus the impossibility 1926, AHQP; Bohr 1927b, 70).
of ascertaining the energy valuesofstationary’states. , Closely connected with the idea of the quantum postulate and the
The concept of complementarity between stationary states and cor- concept ofstationary states was the idea of quantization. The rules of
puscular space-time descriptions, however, seems to presenta serious quantization determined the choice of certain mechanical motions, to
difficulty when large quantum numbers are approached. Accordingto be associated with the stationary states, from a manifold of all pos-
Bohr’s correspondence principle,in the limitof large quantum numbers sible classical motions. In this way a set of integers—quantum num-
the concept of stationary states must approach the classical space-time bers—wasassociated with every stationary state. This choice by dictat-
orbits along which intraatomic particles revolve. This problem occu- ing integers seemed at best arbitrary, at worst a return to Pythagorean
pied Bohr from the time matrix mechanics emerged with its total re- mysticism. Matrix mechanics, which dispensed with the description
hunciation of space-time descriptions. Nor was the problem ignored of stationarystates, also eliminated the idea of quantization.Initially,
by Heisenberg, who though not reluctant to abandon space-time -as I have argued, Heisenberg and Born were quite happyto get rid of
description in the interior of the atom, was nevertheless eager to a postulate of quantization, deducing all the quantum effects, includ-
understand howthe transition from the micro- to the macrodomain ing the.existence of the sequenceof discrete stationary states, from the
might be achieved. Because an electron moving along an orbit with
mathematical formulas of the matrix formalism. The conceptof quan-
a large quantum number should. necessarily be represented by a
tization reappeared, however, in Dirac’s version of quantum mechanics,
wave packet (a superposition of many vibrations), the idea of a sta:
tionary state as a single proper vibration seemed to: Schrédinger ba-
6. When weidentify the exact value of the energy of a stationary state by means of
sically inadequate. How, then, can complementarity, or the mutual collisionsor radiation reaction, we inevitably imply “a gap in the timedescription, which
exclusion of the concepts of Stationary states and individualparticles, is at least of the order of magnitude of the periods associated with transitions between
possibly be maintained in the case of large quantum numbers, where stationarystates. In the limit of high quantum numbersthese periods, however, may be
these ideas are not only no longer contradictory but simultaneously interpreted as periods of revolution. Thus wesee that no causal connection can be ob-
tained between observations leading to the fixation of a stationary state and earlier obser-
applicable?
vations of the behavior of the separate particles in the atom” (Bohr 1927c, 135).
 130 Chapter Six
The Dialogical Birth of Bohr’s Complement
arity 131
indicating that perhaps this idea might have deeper signif
icance than Heisenberg, who were engaged ina controvers
that conceived by matrix physicists. yover the relative merits
Schrédinger’s version of quantum mechanics allowed of wave and matrix mechanics. Schrédinge
the resurrec- rclaimed superiority for his
tion not only ofstationary states but also of the idea of version,due to its greater visualizability, or
quantization. intuitiveness. To counter
According to Bohr,visualization in terms of the Schrédinger, Heisenberg, in his uncer
wave theoretical model tainty paper, presented an ex-
points to the physical interpretation of quantization: “The clusively corpuscular interpretation. Bohr foun
number of d both approaches in-
nodesin the various characteristic vibrations gives a adequate. Bohr agreed with Heisenberg
simple interpre- that Schrédinger’s waves in
tation to the concept of quantum number which was multidimensional space could hardly be consi
already known dered immediat ely in-
from the older methods butatfirst did not seem to appear in tuitive. But neither was Heisenberg’s approachsat
SET ne

the matrix isfactory: the intra-

formulation” (1927c, 126). While matrix mechanics defied atomic reality was not a particle-kinematic
any attempt one. We begin to see the
at physical interpretation, Schrédinger’s wave mechan depthof the gulf between Heisenberg’s and Bohr’s
ics allowed an positions. Only with
understanding of the electric and magnetic properties of atoms. extreme caution, argued Bohr, can one empl
The oy ordinary categories in
wave modelalso allowed an understandingof the failure the interior of the atom; for example, the partic
of the concept le concept may be used
of mechanicalorbits: “From characteristic vibrations with only only during very short times of observation. A
consistent interpretation
a few in such situations is possible only whenthe compatibilit
nodes no wave-packagescan be built which would even approx y of the pos-
imatel y sibilities of observation and definition is closely analyz
represent the motionof.a particle” (Bohr 1928, 113). ed. In general,
Bohr’s discussion of the problem ofinteraction, and especially of the the idea of individualparticles has “just as much oraslittle
“reality’”
problem ofboundparticles in the interior of an atom, is partic as theidea of stationary states.
ularly Not surprisingly, what Schrédinger perceived as the focus
instructive for seeing both the advantages of a wave ontolo of Bohr’s
gy and argument in the Comolecture was Bohr’s defense of the
its limitations. Because of the complementarity of the space-time concept of
and a stationary state. Schrédinger struck back with Bohr’sown
energy-momentum coordinates, for a free particle (and even weapon,
more so arguing that the combination of uncertainty relations and the
for a bound one) precise knowledgeof its momentum and
energy ex- idea of
cludes exact specification ofits space-time coordinates. This stationary states destroys compatibility between definition and
implies the obser-
inadequacy of a particle-kinematic framework and the vation: “It seems to methat there is a very strange relatio
inapplicability n between ©
of the classical concept of mutual forces and potential energy Heisen berg’s uncertainty relation and the claim of discrete
in Heisen- quantum
berg’s approach. This difficulty is “avoided by replacing theclas states. On accountof the formerthelatter can really not
be experimen-
sical tally tested” (Schrédinger to Bohr, 5 May1928, AHOP;
expression of the Hamiltonian bya suitable differential operator” Schrédinger also
(1928, discussed this issue in a letter to Einstein, 30 May
111). However,it is this advantage that showsthelimits of visual
Eee

1928, Przibram 1967).

ization This controversy was not resolved. In the 1950s, Schrod
in the interior of the atom in terms of waves, because the Schréd inger (1952a)
inger criticized Bohr’s idea of stationary states and quantum jumps
equation contains imaginary numbers andis associated exten-
with a multi- sively (see chapter 10).
dimensional space that is “in general greater than the number
of di-
mensions of ordinary space.” Therefore, intraatomic particles,
though Dialogue with Einstein and Compton
adequately represented by a wavetheoretical framew
ROD he "

ork, cannot be
visualized in terms of ordinary space-time pictures.
It is only a three- Another central problem in atomic physicsat the time was
dimensional wave packet of a free particle that can represe that of the
nta particle’s interaction of radiation and matter. More than any other,
space-time location.It is here that the most striking differ this problem
ence between cried out for nonclassical treatment. Nonclassical depart
classical and quantum mechanicslies: while in the former “partic ures had al-
les are ready been made in Max Planck’s and Albert Einstein’s early
endowed with an immediate‘reality, independently of their work, in-
being free cluding Einsteins idea of light quanta (Jammer 1966; Kuhn
or bound” (1928, 114), in quantum mechanics intraat 1978). For
omic, bound par- Bohr the problem of the interaction betweenradiation and
ticles are not visualizable. Consequently, Born’s interp matter led
retation of the to the famous Bohr-Kramers-Slater proposal. When certain
wave function as denoting the probability of position conclu-
should not be sions of this theory were refuted by the Bothe-Geiger experi
applied uncritically to the interior of the atom. ments,
This analysis was addressed simultaneously to Schréd many physicists considered thereality of quanta to have
inger and to been proved
unequivocally (Klein 1970; Stuewer 1974). Accord
ing to the usual
 132 Chapter Six
The Dialogical Birth of Bohr’s Complementarity 133
accounts, Bohr acceptedthefailure of the Bohr-Kramers-Slater theory, was already apparent, according to Bohr, from an analysis of collision
as well as the existence of light quanta, striving from this point on to
phenomena.’
incorporate quanta into his interpretive framework. The principle of
The need to depart from ordinary space-time pictures has a clear
complementarity, announced by Bohr in his Comolecture, was the
meaning in Bohr’s discussionof the interaction of radiation with matter.
culmination of these efforts, resulting in a symmetrical solution of the
Wherever Bohr used the notion of space-time pictures, he discussed
wave-particle dilemma.
models using material particles traversing continuousspatial trajecto-
In whatfollows I argue for a substantial revisionofthis reading. I see
ries. The Bohr-Kramers-Slater theory clearly indicated why such ideas
no evidence of Bohr’s having accepted pointlike quanta.In fact, by us-
could nolonger be applied. In the case of the interaction of light with
ing the de Broglie—Schrédinger idea of a wave packet, Bohr gained a
freeelectrons (the Comptoneffect), the wave picture of light indicated
dramatic insight into past difficulties, enabling him to rehabilitate the
. kinematic peculiarities in the description of the electron’s motion.In the
apparently discredited Bohr-Kramers-Slater theory.
Bohr-Kramers-Slater theory, the notion of the atom as a kinematic sys-
The Bohr-Kramers-Slater paper, a programmatic rather than a tech-
tem oforbiting electrons was replaced bya virtualoscillator model. The
nical work, had mapped out a comprehensive framework for the gen-
visualization of a stationarystate of the atom as a state in which elec-
eral problem of the interaction of radiation with atoms,eliminatin
g the trons movealong well-defined orbits was abandoned(a stepalso indi-
needforthe idea oflight quanta. In this theory, the kinematic model of
cated by Bohr’s analysis of collision processes).’ Bohr’s reaction to the
the atom wasreplaced with a virtualoscillator model.”Thevirtual field
“crucial” Bothe-Geiger experiments. was greater convictionthat the or-
produced by the atom determinedits own probability for spontaneous
dinary space-time concepts of material particles were inadequate, and
emission, as well as the probabilities for the processes of emission and
that a still more far-reaching revision of such concepts wasnecessary.
absorption inducedin other atoms. Such a probabilistic description im-
Perhaps, he argued, de Broglie’s modification of the idea of material
plied not only renunciation of exact conservation laws but—contrary
particles mighthelp to resolve this situation (Bohr 1925).
to the light quanta point of view—independenceof the processes of
This thought, entertained as early as 1925, found its full expression
emission and absorption in atomsthatare far apart.
in the Comolecture. The instantaneousinteraction of light (wave phe-
The conflict between Bohr and Einstein on the nature of radiation
nomenaextended in space and time) with matter (pointlike particles)is
caughtthe attention not onlyof the scientific community but of the gen-
not reconcilable with exact conservation laws—the reasonforintro-
eral public as well (Klein 1970). Theresults of the Bothe-Geiger experi- -
ducingstatistical conservation in the Bohr-Kramers-Slater theory. How-
ments, which were reported in April 1925, did not accord'with the ideas
ever, if the space-time and energy-momentum vectors ofelectrons are
of Bohr-Kramers-Slater. These experiments. detected correlations be-
not sharply defined (as in the case of a wave packet, as opposed to a
tween Comptonrecoil electrons and scattered X-rays, a result more in
localized particle), the wave picture of light can be reconciled with the
keeping with the light quanta conceptof radiation that with the wave
conservation laws. This idea is expressed in all the manuscripts preced-
concept. Most physicists took the experiments as crucial evidence in 4
a ing the Comolecture, as well as in the lectureitself. It is the comple-
favor of the corpuscular idea of light quanta. Pauli, for example, de- 4 mentarity of the space-time and energy-momentum vectors of a wave
clared that from then on light quanta were “as real” as material elec-
packet that makes it possible to unite space-time coordinates and
trons. There is no indication that Bohrever considered the Bothe-Geiger

sek
and Compton-Simon experiments crucial with respect to the nature of 8. “Conclusions concerning a possible corpuscular natureof radiation lack a sufficient
light, even thoughit wasclear to him thatthe program outlined in the basis” (Bohr to Geiger, 21 April 1925, AHQP; BCW, 5:353). A similar position was ex-
Bohr-Kramers-Slater theory could not be retained without modifica pressedin otherletters by Bohr at the'time, as well a8 in the postscript to his “Uber die
seca
- Wirkung”(Bohr 1925), where he attempted to extend the nonconservation ideas of Bohr-
tion. Bohr explained his position in a letter to Hans
Geiger, who had
informed Bohr of the results of his experiments. Bohr further argued Kramers-Slaterto collisions between atoms and materialparticles.
9. Bohr 1925. The misinterpretation that Bohr accepted light quantais related, in my
for the renunciation of ordinary space-time pictures—a necessity
that opinion, to the fact that many authors assume Bohr’s statement about “space-time pic-
7. In each ofits stationary states, an atom continuously emits a tures” applies to the wave propagation oflight as well and consequently concludethat
virtual wave field, the failure of space-time pictures applies to classical wave theory. However, I am not
which is “equivalentto thefield of radiation whichin theclassical theory
would originate aware of anything in Bohr’s writings at the time supportingthis point of view: wherever
from the virtual harmonic oscillators correspondingwith the
variouspossible transitions Bohrused the phrase “failure of space-timepictures,”it referred to the failure of models
tootherstationary states” (Bohr, Kramers, and Slater 1924, 164).
using pointlike material particles.
 134 Chapter Six
The Dialogical Birth of Bohr’s Complementarity 135
conservation principles. What seemed a contradiction is now removed
since according to the character of the description the different aspects
through the introductionof an essential wavefeaturein the description
never manifest themselves simultaneously” (Bohrto Einstein, 13 April
of the behavior of material particles: “The generalcharacter ofthis re-
1927, AHQP; also BCW, 6:420, mytranslation).™
lation [the complementarity of the sharpness of definition of the space-
Bohr’s complementarity viewpoint, formulated in the Como lec-
time and energy-momentum vectors] makesit possible to a certain
ex- furerepresented a “second phase”in his lifelong dialogue with Ein-
tent to reconcile the conservation laws with the space-time coordina
tion stein. This dialogue was, in Pais‘s incisive phrase, “Bohr’s inexhaustible
. of observations, the idea of a coincidence of well-defined events
in source of identity” (1967, 219). Bohr never yielded; his entire episte-
space-time points being replaced by that of unsharply defined
indi- mological edifice was constructed in dialogical response to Eimstein’s
viduals within finite space-time regions” (Bohr 1927d, 93). As opposed
ceaseless challenge.
to the earlier description of the Compton effect, where “in the change
_ of the motion of the electron... one... is dealing with an instantaneous
Dialogue with Campbell
effect taking place at a definite pointin space,” the new wavetheoretical
view of matter indicated that “just as in the case of radiation
. . it is We nowturn to another central problem of the atomic domain—the
impossible to define momentum and energyfor an electron without interaction of matter with matter—andto yet anothercrucial dialogue
considering a finite space-time region” (Bohr 1928, 96). with the lesser known Campbell. Bohr’s first notes for the Comolecture
All these ideas signaled nothing less than the resuscitation of what were triggered by an exchange of letters between Campbell andJor-
seemed’a discredited (Bohr-Kramers-Slater) theory, and a furtherrejec- dan on the interpretation of quantum theory (Campbell 1927; Jordan
tion of the idea oflight quanta. Though Bohr did not argue this point 1927c). Campbell took an early interest in Bohr’s theory andits inter-
explicitly in-his Comolecture, the message is implicit in his analysis. pretation. In 1913, Campbell published a review of Bohr’s riew atomic
WhatBohr only hinted at, Hendrik Kramers,his faithful disciple, chose theory. In the 1920s, Campbell suggested thatthe ideaoforbiting elec-
to announce unambiguouslyin the discussion following Bohr’s lecture. trons in stationary states was a purely formal assumptionand that no
Kramers opened his remarks by saying that he “shall not be able to reality could be ascribed to Bohr’s planetary model. Atthe time, Bohr
add anything fundamentalto Professor Bohr’s exposition.” Kramersin- was not ready to consider such a suggestion—a stand he modified -° -
tended onlyto call the attention of the audience to the examples, which gradually when mechanical visualizationledto crisis of the old quan-
illustrated the resolution of past paradoxes and difficulties with the tum theory. In 1921 Campbell suggested that the paradoxes of quantum
help of the wave theoryof matter: “I am thinking especially of the prin- theory could be avoided by modifying the conceptof time and endow-
ciple of conservation of energy and momentum, which seemed to con-
ing it with statistical significance only. Again, Campbell's idea was not
tradict the wave-theory oflight. ... The difficulty, that the results ofthese greeted with enthusiasm, probably becauseat the time Bohr’s theory
experiments [Bothe-Geiger and Compton-Simon] wereat variance with the wasscoring impressive successes. Encouraged by recent developments
wave theory of light, disappears definitely, if the de Broglie wave theory of that acknowledged the need to modify space-time concepts and to re-
matteris taken into account” (BCW, 6:139, myitalics).1°
nounce visualization (Bohr 1925; Heisenberg 1925), Campbell repeated
Thus the victory of Einstein and his conceptionof light quanta, in the and extended his earlier suggestion thattime betreatedstatistically in
view of Bohr and Kramers, was only apparent. Bohr was more outspo-
ken about this issue in private correspondence and discussions than 11. Similarly, in discussions following Compton’stalk at the fifth Solvay conference,
in formal addresses. As Bohr informed Einstein: “In view of this new held a month after the Como meeting, Bohr continuedto argue againstthe idea of point-
formulation . . . it becomes possible to reconcile the requirementof like light quanta, using explicitly de Broglie~Schrédinger wave concepts. According to
en- Bohr’s description of the scattering process, “we must work with four wave-fields of finite
ergy conservation with the implication of the wave-theory of light,
extension”(two for electrons, before and after the phenomenon, and twofor the inci-
dent and scattered light quanta), whichare localized in the same space-time region. This
10. Kramers repeated Bohr’s considerations, emphasizing that similar
considerations modification of space-time conceptsof particles preserves the ideas of the wave picture
applyto the question ofthe correlationof processes of emission and absorption
in distant of light (and matter) together with conservation laws. Moreover, as Bohr pointed out, in
atoms: while an exact correlation (emission of a light quantum
from the source and ab- the Comptonexperiment“the frequencyshift produced is measured by means of instru-
sorption of the “same”light quantum in the absorbing matter) is out
of the question, an ments, the functioning of which is interpreted according to the wave theory” (BCW, 5:
“approximate” correspondenceis, in fact, maintained, in agreement with
the basic ideas 211). Compton (1927) apparently did not find Bohr’s arguments conclusive, arguing that
of wave mechanics (BCW, 6:139—40).
the last word on this issue had notyet been said.
 136 Chapter Six
The Dialogical Birth of Bohr’s Complementarity 137
the programmatic paper “Time and Chance,” published in
1926. This collision is short compared to the natural periods of the atom . . . the
paper spurred Bohr’s reasoning,asit did Heisenberg’s (chapter 4),
lead- postulate of stationary states would seem to beirreconcilable with any
ing to the formulation of Bohr’s complementarity principle.
Campbell (1926) reminded his readers both of his earlier proposa description of the collision in space and time based on the accepted
l ideas of atomic structure” (Bohr 1926; BCW, 5:851).
that the quantum andclassical theories be reconciled by abandoning
the concept of time in the interior of the atom andof Bohr’s recent This conclusion formed the necessary backgroundto the develop-
dec. ment of matrix mechanics, as well as a starting point for Campbell’s
larations (arrived at “of course quite independently”) that the
descrip- discussion. Campbell's proposal that timebe treated asstatistical in na-
tion of the quantum domain in termsof space and time might be
im- ture wasintendedto provide a way outof this predicament,indicating
possible. Campbell referred to Bohr’s famous postscript to his paper
on how.the space-time description of atomic phenomena must be modi-
the behavior of atomsin collisions, where Bohr took this stand (Bohr
1925). The paperitself was written before Bohr was familiar with fied. Campbell tackled Bohr’s concern with collisions of fast-moving
the particles with atoms, suggesting that if time is statistical, the paradox
results of the Bothe-Geiger and Compton-Simon experiments,
and it resulting from the comparison of “short” collision times with “long”
constituted Bohr’s attempt to provide further guidelines for resolvin
g periods of electron revolution does notarise: “The conclusion thatit
the difficulties of quantum theory by extending the idea of noncon-
[time] is short depends entirely on the assumption that the motionof
servation of energy from the realm of interactions between radiation
the particle andtheoscillation of the atom are uniform. This, of course,
and matter (Bchr-Kramers-Slater theory) into the realm of interactions
wedeny. Particles moving with uniform velocityoroscillating in fixed
between matter and matter (scattering, atomic collisions, ionizatio
n). orbits are undergoing fortuitous transitions betweenthe pointsof their
Bohr’s paperclearly displayed his thinking in termsof concrete models
paths” (1926, 1111).
of atoms where stationary states were represented by mechanically or-
biting electrons, and where processes were categorized as “reciprocal” Campbell'sdiscussion played an importantrole, as I have argued, in
the emergence of Heisenberg’s uncertainty paper. It also provided im-
or “irreciprocal” depending on whether theinteraction timeof particles
passing by the atom is longer or shorter than the natural periods of petus to Bohr’s furtherretreat from the reality of stationary statesin his
Comolecture. However, Bohr could not accept a statistical conception
revolution of the atomic electrons. When an atom collides with an elec-
of time, for it would threaten his overall research program in the atomic
tron of slow velocity, the time of collision is:comparable to the periods
domain, which was intimately connected with classical electromagnetic
of revolution of the intraatomic electrons; so during the time ofcolli-
theory. In this theory periodicity in timeis explicitly assumed. The new
sion, the passage of the atom intoits final stationary state can be com-
quantum theory, advocated by Campbell, clearly “cannot be deduced
pleted and energy is conserved (a’“reciprocal process”). In the case
of from Maxwell's equations in their present form” (Campbell 1926, 1113).
swiftly moving particles (such as a- and f-particles), “the collision
Bohr, who vigorously defended Maxwell's electrodynamics against Ein-
[with the atom] must probably be regarded as finished long before
one stein’s idea of light quanta, and who evenafter the Bothe-Geigerresults
can speak of the completion of a possible transition of the atom from
would only admit a need to modify space-time imagery rather than
onestationary state to another” (Bohr 1925; BCW, 5:198).
Bohr suggested that, as in the case of radiation, so also in the case to revise electromagnetic theory itself, could not agree with Campbell's
of assertion.
collisions of swiftly moving particles with atoms, only statistica
l laws However, Campbell’s focus on the nature of time in dealing with the
of energy conservation apply. This consideration was also
connected space-time problem in the interior of the atom turned outto be very
withthe fact that, while the reaction of an atom upona particle
is ade- suggestive: “The singularposition of time in problems concerned with
quately expressed by continuous parameters, the change
ofstate of the stationary statesis... due to the special nature of such problems” (Bohr
atom is described by discontinuous values. Bohr applied
similar con- 1927c, 131). As for the conflict between space-time descriptions and
siderationsto the case ofcollisions between atoms. The
discussion and conservation laws, discussed by Bohr in 1925 and subsequently treated
the conclusion of nonconservation were firmlybased on
space-time im- by Campbell, the resolution of this conflict constituted another example
agery in the interior of the atom. It was the limitation
of this imagery of the complementarity between space-time andcausality.”
that Bohr acknowledged in the postscript, after he became
familiar with
the results of the Bothe-Geiger experiments implying 12. “If the definition of the energy of the reacting individualsis to be accurate to such
strict conserva-
tion. This point is also stated explicitly in Bohr’s nontechn a degree as to entitle us to speak of conservation of energy during the reaction,it is
ical paper, necessary . . . to coordinate to the transition between twostationary states a time interval
summarizing recent developments: “For impacts in
which the time of long compared to the period associated with this process. This is particularly to be re-
 138 Chapter Six
The Dialogical Birth of Bohr’s Complementarity 139
Clash with Heisenberg: Setting the Historical Record
Straight ambitious Heisenberg withdraw his paper and miss the opportunity to
Weare now better equipped to evaluate the clash makeanother pivotal contribution by letting Bohr reapall the fruits of
between Heisenberg their interpretive struggle. As Heisenberg himself putit: “Perhaps it
and Bohr over Heisenberg’s uncertainty paper,
and to determine the was also a struggle about who did the whole thingfirst’ (interview
effect this clash had on the consolidation of Bohr’s positi
on. An under- withHeisenberg, AHOP).
standing of the wide gap between Bohr’s and Heisenberg’s
interpretive In the postscript to the uncertainty paper, Heisenberg acknowledged
positio ns and professional interests illuminates the intens
e emoti
onal several mistakes in his argument and included Bohr’s objection to his
strain that surrounded the Bohr-Heisenberg dialogue.
The strain was analysis of the Compton effect—Compton recoil applies rigorously
soacute that Bohr and Heisenberg needed to spend
some time away only to free and not to bound electrons. Given this mistake, Bohr must
: from each other, during which Heisenberg wasableto write
his uncer- have thought, Heisenberg had completelyfailed to provide a physical
tainty paper, and Bohrto elucidate his formulation of compl
ementarity. interpretation for the interior of the atom. We haveseen in this chapter
Intense also was the confrontation between Bohr and
Heisenberg over what an ingenious conceptual web of arguments Bohr had spun in
the uncertainty paper: Bohr found the paper mista
ken and premature, order to comprehend “intuitively” intraatomic structure. No wonder
forcefully urging Heisenberg not to rush into print. Heisenberg
did Bohr found Heisenberg’s paper unsatisfactory. Even in thecase of free
notyield, though occasionally “burstinginto tears” as a
result of im- electrons in the Compton effect, as we saw earlier, Bohr found particle-
mense emotional pressure coming from Bohr (inter
view with Heisen- kinematic concepts inadequate. In any type of physical interaction,
berg, AHQP). Heisenberg (1927b) merely agreed to append a
postscript reliance on wave concepts (superposition of waves) was essential,
that acknowledged Bohr’s criticism without revising the conten
t of the accordingto Bohr. Electrons and photons treated as point-masses in
' . paperitself. One of the mistakes Bohr found occur
red in Heisenberg’s Heisenberg’s discussion not only led to technical mistakesin the discus-
analysis of uncertainty during measurementin the y-ray thoug
ht ex- sion; they were “abstractions,” insufficient for the description of real
periment. In his analysis Heisenberg treated both photons and
elec- physical situations. This was the focus of the confrontation between
trons as regular point particles (!), arguing that durin
g their collision Heisenberg and Bohr.'* The issue was not metaphysical preferences
(interaction!) a photon transfers a discrete and uncont
rollable amount (wave-particles as opposed to the mathematical structureof the fonmal-
of energy to the electron (Compton recoil). Yet the Compt
on recoil of ism), as Heisenberg’s recollections would lead usto believe, but Bohr’s
_point-mass particles would not lead to indeterminacy,
as I discussed in dissatisfaction with Heisenberg’s exclusively particle-kinematic inter-
chapter 4, but to precisely calculable changes, as Bohrpointed
out. The pretation. The mistake Heisenberg madein the discussionofthe y-ray
correct explanation, Bohrinsisted, relies on the wavenatureo
flight and experiment, as well as his failure to correct this mistake in the particle
matter in an essential way. I have argued earlier that in oppos
ition to framework, strengthened Bohr’s position. Bohr’s elucidation of the con-
Schrédinger’s accusations, Heisenberg’s uncertainty paper
was aimed ceptof the stationary state, and his resolution of past paradoxesof the
at demonstrating that the “frightfully” abstract matrix mecha
nics is Bohr-Kramers-Slater theory in the wave theoretical framework, made
amenable to a visualizable, intuitive interpretation.
As opposed to the gap between himself and Heisenberg unbridgeable. Bohr’s Como
Schrédinger’s program of wave ontology, aimed at elimin
ating the ma- lecture clearly displays his dissatisfaction with Heisenberg’s position,
trix approach, Heisenberg in his uncertainty
paper used exclusively
particle-kinematic concepts. In this effort, Heisenberg wante thoughina subtle way. Whendiscussingthe uncertaintyprinciple, Bohr
d to avoid pointedto the mistake in Heisenberg’s derivation of uncertainty in the
Schrédinger’s waves altogether (interview with Heisenberg,
AHOP;in- y-ray thought experiment: “Such a change [momentum change during
terview with Klein, AHQP). Clearly, correcting the
mistakes in the way a position measurement] could not prevent us from ascribing accurate
that Bohr suggested, which involved essential relian
ce on wave con- values to the space-time coordinates, as well as to the momentum-
cepts, would undermine Heisenberg’s aims.
Nor would the intensely
energy components before and after the process. The reciprocal uncer-
membered when considering the passage of swiftly
moving particles through an atom.
According to the ordinary kinematics theeffective durati 13. “I argue with Bohr over the extent to which therelation p,q, ~ h has its origin in
on of such a passage would be the wave—orthe discontinuity aspect of quantum mechanics. Bohr emphasizes that in
very small as compared with the natural periods of
the atom, andit seemed impossible
to reconcile the principle of conservation of energy the gamma-ray microscope the diffraction of the wavesis essential; I emphasize that the
with the assumption of the stability
of stationary states” (Bohr 1928, 116). theory of light quanta and even the Geiger-Bothe experiments are essential” (Heisenberg
to Pauli, 4 April 1927, PC).
 140 Chapter Six
The Dialogical Birth of Bohr’s Complementarity 141
tainty . . . is an outcomeof the limited accuracy with which changesin
example of the correspondence,” Bohr presented matrix mechanics as
energy and momentum can be defined, provided the wave-fields used
the culmination of his own research program, based on the correspon-
for the determination of the space-time coordinates of the particle shall
dence principle: “It is only through the quantum theoretical methods
be sufficiently limited” (1927c, 120-21 ).
created in the last few years that the general endeavors laid downin the
Some contemporary physicists understood well the point Bohr was
[correspondence] principle ... have obtained an adequate formulation”
making. Ehrenfest, in a remarkable letter to Goudsmit, Uhlenbeck,
and (1927c, 124). Bohralso expressed his reservation about theinitial em-
Dieke, referred to Bohr’s derivation of uncertainty, relying on “wave
phasis on observability in papers on matrix mechanics, finding this
kinematics” and “amending the error running through the Heisenberg
emphasis not only historically misleading but physically inadequate.
paper” (der durchlaufenden Fehler von Heisenberg; Ehrenfest to.Goud
- Only in “a‘certain sense” may matrix mechanics be described as a “cal-
smit, Uhlenbeck, and Dieke, 3 November 1927, AHOP, translated in
BCW, 6: 37-41, quote on 39). Born, in a discussion following the Como culus with directly observable quantities,” becauseit is “limited just to
© problems, in which in applying the quantum postulate the space-time
lecture, accepted Bohr’s derivation of uncertainty as following, not from
‘description may largely be disregarded, and the proper question of ob-
uncontrollable changes during measurement, but from the wave na-
ture of matter (BCW, 6:137-38). If, for Heisenberg, uncertaintyfollo
servation therefore placed in the background” (1927c, 125). The corre-
wed spondenceprinciple, according to Bohr, continued to serve as a neces-
from discontinuity (discontinuous changes), for Bohr, uncertainty fol-
sarybasis for the interpretive framework of physics, together with the
lowed from a dialectical combination of continuity and discontinuity,
indispensability of classical concepts. This is the reason for the amal-
or “individuality and superposition.” For Bohr, at thispoint, not the - - gamation of “historical” and “philosophical” aspects in the Comolec-
operational definitionof concepts, but agreement between the possibili-
ture, which has puzzled some scholars of Bohr’s work (Folse 1985; Gib-
ties of definition and observation was essential.
bins 1987). Bohr’s subtle message registered fully with Heisenberg, who
The Comolecture contains other, more subtle hints of Bohr's past dis-
from that timeon,in his historical expositions, presentedhis reinterpre-
agreements with Heisenberg. The clash over uncertainty wasnotthefirst
tation paperas following from Bohr’s correspondenceprinciple.
time Bohr hadreasonto be dissatisfied with Heisenberg’s position. As
Bohr’ssecretary, Betty Schultz, recalls (interview with Schultz, AHOQP),
. Confrontation with Pauli
Heisenberg was not very helpful.to.Bohr, certainly not as useful as the
ever loyal Kramersor, later, the agreeable Klein (about Kramers’s cru-
According to the usual historical accounts, the confrontation between
cial role in Bohr’s research program, see Dresden 1987). Nor was Hei-
Heisenberg and Bohr ended when each comprehendedtheother's point
senberg always scrupulous about acknowledging his debt to Bohr. In
of view, realizing that there was no real contradiction betweentheir
hisreinterpretation paper (1925), Heisenberg does not cite Bohr’s work
positions. But the text of the Comolecture, as well as Bohr’s and Hei-
atall, despite thefact that the paper wasbuilt on Bohr’s correspondence
senberg’s scientific correspondence at the time, reveals the continuing
principle in a fundamental way. Nordoes Heisenberg mention Bohr’s
gap between their interpretations of the atomic interior. For Bohr, the
conclusion from his 1925 paper aboutgiving up space-timevisualiza- Heisenberg-Born-Pauli statistical kinematic interpretation of particles
tion inside the atom. Instead, Heisenberg presented his work as flowing
(“abstractions”) was not sufficient for a consistent interpretation of
from thepositivist principle of elimination of unobservables. I have ar-
quantum physics; the wave nature of free “individuals”(electrons, pho-
gued that this was not a guiding principle, buta justification after the
tons) hadto betakeninto account in an essential way. The uncertainty
fact, and that the conceptual package— correspondenceprinciple,
non- relations, according to Bohr, apply rigorously only to free particles (as
conservation, wave theory of light—seemed to be discredited
by the a direct consequenceof the limitation of the associated wavefields) and
Bothe-Geigerresults. It is likely that Heisenberg chose not to mention
the correspondence principle for strategic reasons. Bohr must
in generalcases of interaction “must always be applied with caution”
have been (Bohr to Schrédinger, 23 May 1928, AHOP, translated in BCW, 6:49). In
unhappy aboutthis, as well as about the fact that Heisenberg’s
matrix particular, the uncertainty relation between position and momentum
mechanics suppressed the idea of quantization and “swallowed”
the for discrete stationary states of an atom cannot hold true without modi-
idea of stationary states.
In his Comolecture Bohr set the historical record straight. Aftercit fication. According to Bohr, the application of the concept ofa station-
- ary state and the tracking of the behavior of individual particles in the
ing work by Kramers and Rudolf Ladenburg as
a “characteristic atom are mutually exclusive.
 142 Chapter Six
The Dialogical Birth of Bohr’s Complementarity 143
As oneof the main architects of the particle-kinematic interpre
tation Conclusion
of the interior of the atom, Pauli did not agree with Bohr. For Pauli
it
was meaningful to talk about the uncertainty relation for
particles in The united public front did not imply that the Copenhageninterpreta-
the case of individual discrete Stationary states, as was discuss
ed in tion was coherentor consistent. The “interpretation” was actually an
Heisenberg’s paper. Nor did Pauli agree with Bohrthata visualiz
able, amalgamation of the different views of Bohr, Born, Heisenberg,
intuitive interpretation is possible only in three dimensions, arguing Pauli,
and Dirac. Thereis no indication that Heisenberg fully accepted Bohr’s
that his own interpretation of the wave function could “in principle
be views. Unlike Bohr,the “mathematical physicist” Heisenberg preferred
empirically ascertained by statistical utilization of results of observa
- one coherent set of concepts, rather than two incompatible ones." In
_ tion” (Pauli to Bohr, 17 October 1927, PC ). The Bohr-Pa
uli dialogue had this respect, Heisenberg was muchcloser to the “enemy” Schrodinger,
a crucial effect on further elaborations of complementarity.
Pauli, an whoconsidered Bohr’s solution merely a judicious escape. Heisenberg’s
early critic of continuousfield theories, was not as enthusiastic as Bohr
position was that wave language andparticle language, being equiva-
aboutthe indispensability of classical electrodynamics. He did accept
lent descriptions of the same reality, were mutually convertible. One
the Bothe-Geiger results as conclusive evidence of the physical reality
could therefore use either languageat will, that of particles or thatof
of Einsteinian light quanta. Bohr’s complementarity of space-time and waves, without needing to use them simultaneously, as Bohr claimed
causality was not alwaysclear to Pauli (Pauli to Bohr, 13 January 1928,
(interview with Heisenberg, AHOP).
PC). The.first comprehensive discussion of wave-particle duality, as far According to the usual accounts, soon after their heated arguments .
as I know,is presented in Pauli’s 1933 encyclopediaarticle. Pauli, then, over the uncertainty paper, Bohr and Heisenberg reached complete
was mostlikely the architect of the symmetrical complementarity of agreement. Yet genuine unanimity of opinion between the two men ~~
wavesandparticles, as distinct from Bohr’s different notion of the
com- never occurred. Rather they realized that “all that mattered now wasto.
plementarity of space-time and causal descriptions. oo presentthe facts in such a way that despite their novelty they could be
Bohr’s and Pauli’s versions of the interior of the atom were not, as grasped and acceptedbyall physicists” (Heisenberg 1971, 79). The need
far-as I know, ever reconciled. Subsequent elaborations of complemen- to offer a unified explanation, capable of countering the opposition, was ;
tarity (two-slit experiments, interaction of atomic objects with measur
- one ofthe reasons for the obscurity of Bohr’s Comolecture, where the
ing devices) centered on individualparticles. Bohr and Pauli did, how- differences between Bohr’s and Heisenberg’s positions were subdued.
ever, agree entirely on the importance of elucidating the concept of Such obscurity was necessary in order to conceal the conceptual gap
measurementfor the interpretation of quantum physics (Pauli to Bohr, between their interpretations of physics for the interior of the atom.
17 October1927, PC). Subsequently, the Nature version of the Comotalk The uncertainty principle for the position and momentum of a particle
(Bohr 1928), which was conceived in dialogue with Pauli, placed heav-
cannoteasily be reconciled with the representation of separate station-
ier emphasis on measurementthan had the original version delivered ary states by harmonic partial vibrations. This difficulty was oneof
at the conference, written in collaboration with Klein and Darwin.If the
the early targets for Schrédinger’s criticism. Einstein agreedfully with
Comolecture stressed harmony betweenthe possibilities of definition Schrédinger that the particle-kinematic framework, even when sup-
and observation, the later Nature version ushers in the key theme
of plemented by the uncertainty principle, was deficient, and that the
the “uncontrollable element” introduced by observation. Bohr’s and “shaky” concepts p and q should be abandoned: “The whole thing was
Heisenberg’s positions seem to have becomecloser. In addition
, Bohr’s invented forfree particles and suits only this case in the natural way.”It
enthusiastic praise of Schrédinger’s theory was subdued in the Nature
wasnot becauseof their “classical nostalgia” that Einstein and Schré-
version. Bohr’s claim in the Comolecture that “just Schrédinger’s for-
dinger were notpersuaded by the ‘“Bohr-Heisenbergtranquilizing phi-
mulation of the problem of interaction seems particularly well-suited losophy” (Einstein to Schrédinger, 31 May 1928, Przibram 1967).
for theillustration of the nature of the quantum theory” waseliminated
Yearslater, Heisenberg conceded that Bohr’s matter wave interpreta-
from the Nature article. Similarly eliminated was the claim about the
tion for the interior of the atom was perhaps “muchcloser to the truth”
utility of Schrédinger’s wave mechanics for a “demonstration of the
consistency of symbolic [matrix] methods” (1927c, 128). With 14. “If one immediately starts with the supposition there are both waves and particles,
Pauli’s
skillful assistance, the G6ttingen-Copenhagen front was uniting and everything can be madecontradictory free,” Heisenberg wrote disapprovingly to Pauli
consolidatingits stand. two months after the uncertainty paper appeared (May 1927, PC). See chapter11 for fur-
ther discussionofthis issue.
 144 Chapter Six

than his own (1958, 51). Yet the conflict between Bohr and Heisenberg
on this issue had been circumvented rather than resolved. Subsequent
CHAPTER 7
interpretive developments shifted the emphasis to the measurement EX
problem,the algebra of Hilbert spaces, quantum logic—areas of in-
quiry dissociated from the initial struggles over what happensinside
The Challenge of Einstein-Podolsky-Rosen
the atom. As Heisenberg revealed, the physicist had no choice but “to
withdrawinto the mathematical scheme” when the vague useofclas- and the Two Voices of Bohr’s Response
sical concepts, encouraged by Bohr’s complementarity, led to difficul-
ties and inconsistencies (1958, 51). This sort of escape was not-open to
While imagining that I understand the position of Einstein, as regards the
Bohr—a “natural philosopher” rather than a “mathematical physicist”: 4]
EPRcorrelations, I havelittle understanding of the position of his principal
(see chapter 12). To the end of his life Bohr struggled to clarify and opponent, Bohr. Yet most contemporary theorists have the impression that
extend his complementarity framework. In order to comprehendfur- Bohrgot the better of Einstein in the argument and are under the
ther developments, as well as the numerous contradictions in the ortho- impression that they themselves share Bohr’s view.
doxinterpretation of quantum physics, a historical perspective on the
initial interpretive efforts is necessary. J. S. Bell 1987, 155
Noris the genesis of Bohr’s complementarityprinciple fully compre-
hensible without taking into account such psychosocialfactors as am-
bition, professionalinterest, group dynamics. It is remarkable, though
not surprising, how much the cognitive positions of different contribu-
tors coincided with their professional and personal interests. ‘Thus TwoVoices in Bohr’s Response to Einstein-Podolsky-Rosen
Bohr, in hisComolecture, emphasized the wave aspect of matter.and
light;the wave ontology allowed him to preserve and entrenchhis ma- The strongest challenge ever posed to the orthodox philosophyof quan-
jor contributions to science—the idea of stationary states, the approach tum physicsis the Einstein-Podolsky-Rosen argument (Einstein, Podol-
of the Bohr-Kramers-Slater theory, the statistical description of atomic sky, and Rosen 1935, hereafter EPR).! The argument,if one accepts some
collisions. Heisenberg, who had no direct investmentin these ideas, intuitive and, at the time, widely accepted notions of physical descrip-
was readyto dispense with the wave aspect altogether, in order to tion (reality), pointed out that something peculiar, and perhapsunac-
argue the superiority of matrix mechanics, to which he had contrib- ceptable, is implied by the quantum formalism: Quantum theory is
utedso decisively. Emotional intensity in a scientific dialogue is not an either incomplete, or inconsistent, or both (Beller and Fine 1994). In
aberration; it is vital fuel for the shaping of ideas, the formation of particular, the conjunction of the completeness of quantum theory and
stands, the achievementof breakthroughs. In dialogical accounts that the separability of states of distant systems cannot be maintained. The
acknowledgetheessential formativerole of scientific controversies, the argument came as a “bolt from the blue,” and its effect was “remark-
line betweenthe “cognitive” and the “social” becomes blurred. - able” not onlyon Bohr but also on other quantum physicists (Rosen-
Equally problematic becomes the idea of a definable research pro- feld 1967, 128-29). Diracinitially considered the argument devastating:
gram. A simplistic division between “orthodox” and “opposition”is “Paul Dirac said: ‘Now wehaveto startall over again, because Einstein
not adequate for describing a living, creative stage in the formation of proved that it does not work’” (interview with Bohr, 17 November
ideas. Wesee, rather, flexibility, simultaneous openness to 1962, AHQP). —
different
conceptual options, at times genuine dialogue, at times “infiltration” Bohr's pre-1935 philosophy contained as an intrinsic part the concept
or selective appropriation of opponents’ ideas. We saw that on of a robust physical disturbance. The challenge of EPR undercut Bohr’s
certain
points Bohr wascloser to Schrédinger than to Heisenberg, while on idea of a direct physical disturbance. Einstein, Podolsky, and Rosen
other issues Heisenberg was closer to Schrédinger than to Bohr. At discussed a system of twoparticles that interacted initially and then
times, emotions becomefierce, but no concessions are made; movedapart.If one directly measures the value of either of two conju-
this is
whathappensat the crossroads of vital interests:
1. My presentation of Bohr's response to EPR (Bohr 1935a) follows Beller
and Fine
(1994) closely. I am deeply grateful to Arthur Fine for most rewardin
g worktogether.
RATE
146 Chapter Seven
Einstein-Podolsky-Rosen and the Voices of Bohr's
Response 147
gate variables for one system, one can predict with certainty the
Q, ~ Q2 commutes with P, + P,, we can by
(unmeasured) value of the same quantity in the other system.” The au- measuring Q, calculate Oo;
or by measuring P, calculate P,. Whatis the reaso
thors of EPR investigated the state function of a two-particle system n that contra Einstein,
wecannotdo both simultaneously?
and its “reduction” to the state functions of the component systems
Let us consider Bohy’s first answer — the operationa
during measurement. They reached the puzzling conclusion thatas l one:5 In addition
a - to the first two-slit diaphragm, we employ a second
consequence of two different measurements, “without in
any way dis- diaphragm sus-
pended by weakspringsorrigidly bolted, depen
turbing” the second, distant system, twodifferent function
s can be as-
ding on whether we
intend to measurethe position or momentum ofthef
signed “to the samereality.” The unmeasured particle has a reality that. irst particle (see
diagram on next page). We can measureP,of thefirst
is simultaneously describable by an eigenfunction of position and: particle by using
an. the second diaphragm, and deduce P,—the momentum of
eigenfunction of momentum—using a standard eigenstate-eigenvalue the second -
particle. For the measurement of P, we use the movabl
rule, the authors concluded that both position and moment e diaphragm, so
um can, we exclude in principle the possibility of measuring Q,, and
with certainty, be ascribed to the second particle. As quantum me- thus we
exclude in principle the possibility of predicting Q.—the positio
chanics forbids assignmentof definite values to two conjugate variables n of
the secondparticle. This answer accords well with Bohr’s stateme
simultaneously, the authors concluded that quantum mechanics was nt that
measurements onthefirst particle imply “an influence on the very con-
incomplete.*
ditions which define the possible types of predictions regarding thefu-
During the course of their argumentation, the authors proposed a
ture behavior of the system” (1935a, 700).¢ If, operationally, we equate
“criterion of reality”: “If, without in any way disturbing a system, we
measurement, and associated predictions, with definability, we see that
can predict with certainty the value of a physical quantity, then there —
there is no possibility of simultaneous prediction, and therefore
exists an element of physical reality corresponding to this physical defin-
ability, of the position and momentum of the second particle. - -
quantity” (EPR, 778).4 Bohr did not challenge their definition, finding
This answerisboth too short and too long. Too short, because there
only an “ambiguity” in the expression “without in any way disturbing
are many sentencesin Bohr’s text (in additionto his repetition of
a system.” Bohr’s answer to EPR centered on the idea of disturbance. some
“simple, and in substance well-known,” examplesof measurementar
Twodifferent, even incompatible, answers are concurrently present in -
rangements) that seem to be either superfluousorirrelevant (they: find,
Bohr’s response to EPR. One is tempted to assumethat Bohr must have
as I will shortly demonstrate, a natural place from the perspec
presented only one unequivocal answer and the difficulty is that of the tive of
Bohr’s second answer). Yet the first answer is also too long, becaus
reader. Yet such an assumption renders muchof Bohr’s text superfluous e the
operational reply does not require the “physical actualization’”of the
or incomprehensible. The two voices in the paperarereal, with one
EPR argument—theoperationalreplyis a general one, indepen
voice rooted in the past (before 1935), the other emerginginto the future dentof
the specifics of the experimental arrangement. Indeed, when Bohr sum-
(after 1935).
marized his response to EPRin 1949, instead of twoparticles, he consid-
Bohr’s response (or responses) to EPR wasbuilt on the “physical ac- ered “for the two parts of the system... a particle anda diaphr
tualization” of EPR’s mathematical reasoning, rather than dealing agm.”
di- The operational answeris thus compressed into a few lines: “After
rectly with the consequences of the formalism. Bohr proposed for the the
particle has passed through the diaphragm, we have in principle the
measurement arrangement a diaphragm with twoparallelslits through
choice of measuringeither the position of the diaphragm or its momen
which bothparticles pass simultaneously. If such a diaphragm is sus- -
tum and, in each case, to make predictions as to subseq
&& pended by weak springs, we can know Q, — Q, and P, + P,. Because uent observa-
E tions pertaining to the particle. As repeatedly stressed, the princip
& al
§ 2. In the EPR mathematical presentation, the variables are a position coordinate
and
pointis here that such measurements demand mutually exclusive ex-
& linear momentum. : perimental arrangements” (Bohr 1949, 233). A simultaneous reality,
3. The argumentof EPR is a complex one— it was first analyzedin ac-
Fine (1986, chap. 3) cording to the operational approach, can be ascribed to two
SF
and extended in Beller and Fine (1994). In particular, Beller and Fine
discussthe issues of
variables
“incompleteness”and “inconsistency” in the EPR argument, and the crucial difference onlyif they are simultaneously measured.
s
between the EPR argument and Bohr’s summaryof it.
4. The authors did not use this criterion in the paper. The onlytimet 5. It is this answerthat Bohr himself later singled out (Bohr 1949);
heyreferred to it most commentators
was to demonstrate that such a criterion is consistent with the eigenstate follow Bohr’s later presentations, unaware of the second voice,
discussed in Beller and
-eigenvalue rule, Fine (1994).
whoseapplications constitute accepted “quantum mechanicalideasof
reality” (EPR, 778; 6. It was here that Bohr found “an ambiguity as regards the meaningo
see also Beller and Fine 1994). fthe expression
. ‘withoutin any way disturbing the system.’””
raereeste
 Einstein-Podolsky-Rosen and the Voices of Bohr’s Response 149

Let us note that this simple answer was provided before Bohr by
Arthur E. Ruark (1935).” His crisp reply does not leave much ground
for the Bohr-Einstein dialogue—the opponents can only acknowledge
a fundamental difference between their metaphysical presuppositions.
Einstein envisaged such a simple (and for him, deeply unsatisfactory)
countermovecalling such a reply “unreasonable” (EPR, 780). Bohr’s
aim was not confined to amicably stating the metaphysical disagree-
ments between Einstein and himself—Bohr wanted to “convince” Ein-
stein, and to emerge “a victor” from this confrontation. If the opera~
tional reply is indeed what Bohr offered, it is not clear why he needed
six weeks of “utmost concentration and unrelenting efforts” to fash-
ion such a response (which was suggested, before Bohr, by Ruark and
even Einstein himself!). Where in the operational reading do wefind
the “painstaking scrutiny of every detail” (Rosenfeld 1967, 131)? There
clearly must be something more in the text than this simple operational
countermove._
Anotherreading of Bohr’s reply finds his second answer, which deals
with the “ambiguity” in the concept of disturbance in a moredirect
-way (Beller and Fine 1994).It is this reading that makes good sense
of Bohr’s discussion of the details of the specific physical actualization
that he proposes(details that do not make sense under the operational
reading).
The measurement of momentum is similar to that outlined in thefirst
answer—we measure P, + P, with the help of the two-slit diaphragm
(through which both particles pass simultaneously). The two-slit dia-
phragm must be movable (suspended by weak springs orthe like). We
can subsequently measure the momentum ofthefirst particle, using a
second movable diaphragm,spatially separated from the first. Know-
ing P; + P,,and measuring P,, we can easily predict P, (becauseof the
law of conservation of momentum, the two diaphragms can indeed be
at any distance from each other).
The position measurement, however, presents a totally different con-
ceptualsituation. In fact, Bohr’s physical actualization of the position
measurement violates the EPRcase. In Bohr’s setup, as opposedto the
EPR setup, Q; — Q, has definite value atone instant only——whenthe
twoparticles pass throughthe twoslits of the first diaphragm. At any
other point, the value Q; — Q, becomes indefinite, according to Schré-
Bohr’s analysesof simple thought experiments in terms ofclassical concepts
dinger’s equation. Yet this means that if we use a second diaphragm,
werecentral to the
rhetoricof the “inevitability”of the Copenhagen interpretation. These analyses
spatially separated from thefirst (the two-slit diaphragm), to measure
created theillusion
that no knowledgeof the mathematical formalism of quantum mechanics
is needed in order to com-
Q,, we cannotpredict Q,! (We have lost our knowledge of Q,; — Q,.)In
prehend its philosophicallessons. . 7. “This conclusion [EPR‘s] can be attacked by anyone whoprefers to say that P
Therigidly bolted apparatus (center) is for position measurement. The apparatus and
suspended by Q could possessreality onlyif [they both] could be simultaneously measured”
weak springs(fop) is for momentum measurement Diagrams adapted from (Ruark
Bohr (1949, 216, 219, 220). 1935, 466).
8. The possibility of such a reading wasfirst suggested to me by Alon Drory.
 150 Chapter Seven
Einstein-Podolsky-Rosen and the Voices of Bohr’s Response 151
order to be able to predict Q, from the measuremen
t of Q,, we must this notion to a dead end. The new, emerging operational voice will
measure Q, at the very moment the two particles pass throug
h the culminate in unreserved verificationism and a future repudiation of the
two-slit diaphragm! This meanseither that the second diaphr
agm must notion of disturbance. Such complex, polyphonic explorationis fitting
be infinitely close to the first or that the two diaphragms(the two-sli
t for the interface between the old and the new —thetransition to a sub-
one and the one for measurementof Q,) actually mergeinto a single
stantially different position is not a Gestalt switch.
arrangement.
Othertensions, consequently, inhabit Bohr’s paper, such asthe ten-
This reading accords well with certain passages in Bohr’s reply to
sion between “consistency” and “inevitability” arguments: thefirst, op-
EPR, which seem to be strangely out of place under an operational
erational answer introduces the inevitability motive, while the second
reading. For example, if we measure the position Q, ofthefirst
particle , answer is more compatible with Bohr’s assertions of “soundness,” “lack ©
welose the possibility of knowing P, + P5. This loss of. knowledge oc-
of contradictions,” “rationality,” and “consistency”of the quantum me-.
curs only because the common diaphragm (whichincludesthe two-sli
t chanical scheme (see chapter 9). Similarly, Bohr strikes two different
one for the measurement of P,; + P;) must now be rigidly bolted
and chords on the themeof “the radical revision of ourattitude towardsthe
cannot serve for momentum measurement: “We have bythis
procedure concept of physical reality” (1935a, 697). The first, operational answer
cut ourselvesoff from any future possibility of applying the lawof con- forbids any talk about objective, observer-independentreality. This an- :
servation of momentum to the system consisting of the diaphragm and swer will culminate in Bohr’s redefinition of physical phenomena—no'*
the two particles”” (1935a, 697). There are other sentences that seem to elementary phenomenonis a phenomenon until it is an observed phe-
make sense only under the specific condition of merged diaphragms nomenon.In relation to the revision of the conceptof reality, the second
(Beller and Fine 1994, 14). answer doesnotcontain anything new—Bohr merely repeatshis pre-
Clearly, the realization of position measurementintroducesrestric-
1935 argumentsof “inseparability” of measuring and measured, of ob-
tions and physical effects not present in the EPR paper. In the EPR case ject and subject (such inseparability necessitates a change in the concept
both Q; — Q, and P, + P, are determinable simultaneously with either ofreality, as I discuss below). This reply accords well with Bohr’s asser-
the position Q, or the momentum P, of the first particle, while Bohr’s
tion that his discussion of the EPR case does not contain any greater
double-slit arrangementdoesnotsatisfy this requirement. In Bohr’s ex- intricacies than those in his previous analyses: of simple mechanicalar-
ample only one of Q; — Q, or P, + P, could be determined simulta
ne- rangements discussedin his publications before EPR.
ously with the variable one chooses to measure onthefirst particle.
In Our description of the conflicting voices in Bohr’s reply discloses
addition, the EPR condition that “no real changecan take place in the
why Bohr’s readers have such difficulty following his arguments.’ Un-
second system in consequence of anything that may be donetothefirst less we acknowledgethe different voices in Bohr’s paper, its tensions
system”is violated in Bohr’s arrangement. Bohr’s physical realization and contradictions, parts of the paper are incomprehensible. Nor does
contains an indirect mechanical disturbance, because changing the the great labor spenton it, and “the state of exaltation in which Bohr
measurementfrom Q, to P, of the first particle demandsa changein the
accomplished this work” (Rosenfeld 1967, 131) make sense.
mechanical arrangement of the two-slit diaphragm, through which
both particles pass (and, consequently, in the second diaphragm, which
Bohr’s Victory?
mustbe attached to thefirst).
We can see now why Bohr does not challenge the EPR conception of: There is a widespread myth that Bohr enjoyed a triumphoverEinstein -
reality and only finds an “ambiguity” in the expression “withoutin any
in their dialogue on EPR(see Rosenfeld 1967; Wheeler and Zurek 1983,
way disturbing a system.” A real physical disturbance does exist in 142-43). Yet none of Bohr's answers are satisfactory. I have already
Bohr's (incorrect) realization of the case proposed by EPR! Bohr’s
sec- pointed out that the answerbuilt on the concept of disturbance is fun-
ond reply is a veritable failure. No wonder Bohr never repeated this
damentally flawed. Theother, operational answer can be more proper-
intricate, yet fundamentally flawed argument. Refusing (or unable) to
ly seen as an enforced, ad hoc legitimation move, rather than as an
explore quantum “wholeness” in terms of quantum ontology, Bohr’s
only choice wasto land in positivism.
Twodifferent voices, then, meet (or clash) in Bohr’s respons 9. Bohr himself wrotelater that he was “deeply aware ofthe inefficiency of expression
eto EPR. which must have madeit very difficult to appreciate the trendof the argumentation aim-
The old voice, holding on to the notion of physical disturbance, brings
ing to bring out the essential ambiguity” (1949, 234).
152 Chapter Seven
Einstein-Podolsky-Rosen and the Voices of Bohr’s Response 153

adequate confrontation with the deep challenge of the EPR correlations. Westarted with two superimposed answers to EPR by Bohr. We end
In particular, Bohr’s operational answer conceals rather than reveals en- up with no answeratall! It is bewildering that Bohr’s response wasever
tanglement and nonlocality, which is the most powerful message of considered, and is often still considered, an adequate (not to mention
EPR.” Even if we accept Bohr’s changingthe rules of the game, and his “triumphant”!) reply to EPR. I can suggest a few explanations for this
refusal to enter into ontological inquiry, his operational reply falls apart strange state of affairs. The myth is in part connected with the general
underclose scrutiny. Quite aside from the general weaknesses ofthe mythology of the Copenhageninterpretation, the hero worship of Bohr,
operational positivist stand," operationalism is especially unsuited— . - the fabrication of the “winner's narrative’”—issuesI discuss in the chap-
in fact it undermines —Bohr’s philosophy of complementarity. ters that follow. Yet we also have to take into account the specifics of
In his discussionof a position measurement whenthe diaphragm is . _ the Bohr-Einstein confrontation over EPR. A few ingenious rhetorical
rigidly bolted, Bohr arguesthat an “uncontrollable” amount of momen- moves characterize Bohr’s response and create the illusion of victory. |
tum “passes into . . . support” (1935a, 697). The momentimm transferred '. By giving a short, nonmathematical summary of the dense and complex
from theparticle to the measuring device cannot be measuredin prin- EPR paper, Bohr ensured that few would bother to read the EPR paper
ciple.If it cannot be measured from an operational pointof view,it has 3 itself. (Those who did, and immersed themselves in the details of the
no meaning. Yet in this case, quantum uncertainty, contra Bohr, cannot EPR argumentation without bias, obtained a rich harvest: Bell 1964,
arise (still less “inevitably” follow) from the physical interaction be- . 1966; Bohm 1952.)
tween an object and a measuring device (Beller and Fine 1994, 21).!* The _ The overwhelming majority. of presentations of the Bohr-Einstein de-
conclusion of Fine and myself wasthat = ~~ bate use Bohr’s nonmathematical summary of EPR. Yet this summary is.
‘misleading and introduces weaknessesnot present in the original EPR
paper: Those wholightly dismiss the EPR challenge (Pais 1991, 429-31)
from the positivist perspective that Bohr eventually adopted,the idea
dismiss in fact Bohr’s version ofit (see Beller and Fine 1994, 2—6, for a
of an uncontrollable exchange of momentum, which is supposed to
ground his physical picture of quantum uncertainty, is problematic. full discussion of the difference between EPR and Bohr’s EPR). A par-
The only way around the problem seemsto be to turn the ground up- ticularly obvious weaknessis Bohr's assignment of simultaneous posi-
side down, and to make the measurable uncertainty the operational tion and momentum values to the unmeasured particle. According to
basis for.the language of uncontrollable exchange. Thus despite the Bohr’s summary, if we measure the Q,, then, according to the criterion
lively imagery, whenBohrtalks of an exchange or transfer of momen- of reality, we can assign a definite value to Q,; similarly, if we measure
tum,there is literally nothing (and in particular, no momentum)that P,, then, accordingto this criterion, we can assign a definite value to P,.
is transferred or exchanged. Bohr conjures up a robust physical pic- Yet, accordingto the criterionofreality,it is not clear why we can assign
ture: the feature of wholeness or “individuality” of the quantum phe- Q. and P, at the same time, unless we had measured Q, and P, simul-
nomena connected to an uncontrollable interaction between object and taneously. The weaknessis so obvious that it seems we can dismiss the
apparatus—all giving rise to the quantum uncertainty. Uponscrutiny,
EPR argumentat once. Consequently, we might conclude, there is no
however,this impression turns outto be the effect of a conjuring trick.
Only the quantum uncertainty itself is independently meaningful.
reason to enter into the mathematical intricacies of EPR argumentation.
From the positivist pointof view,the rest is a word-picture constructed As Fine and I have argued, EPR’s demonstration of simultaneous P
around the experimentally verifiable uncertainty formulas, like a col- and Q values depends noton thecriterion of reality but on the state
lage of printed wordsglued on to a radiantobject. (1994, 22-23) descriptions in accord with the eigenstate-eigenvalue rule (the criterion
of reality is only introduced to showthatit is consistent with this rule;
10. As argued in Beller and Fine (1994), Bohr’s talk about the “wholeness of the experi- Beller and Fine 1994). The main point, and the strength of the EPR
mental situation” reflected his positivist solution to EPR (where the operational presup-
position implies the inclusion ofall the aspects of conditions of measurement), and not
argument, was to challenge the adequacy of the quantum mechanical
his endorsementof a nonlocal or nonseparable conception of reality. In fact, Bohr, as did characterization of a system’s state by means of a wave function. A
Einstein, considered the option of nonlocality unacceptable (Beller and Fine 1994). Bohr’s mathematical elucidation in the EPR paper demonstrated that such a
notorious ambiguity creates room for later, charitable scholars to ascribe to him insights description introduces nonclassical features of entanglement, or corre-
aboutthe nonlocality of the EPR situation that he in fact did not express (see chapter12). lation, that are at odds with the deeply entrenched intuitions about the
11. I discussed these weaknesses in chapter 3.] will further analyze this issue in chap-
individuation of physical systems. Bohr’s presentation of EPR conceals
ter 8 and argue that positivism is not a natural stand for the working scientist.
12. Beller and Fine discuss some potential responses to this difficulty and find them this crucial insight.
unsatisfactory (1994, 21). Thus the notion of “physical reality” is the focal point of Bohr’s
 154 Chapter Seven
Einstein-Podolsky-Rosen and the Voices of Bohr’s Response 155
summary of EPR, but not of the EPR paper itself. Bohr, as we have
As I have mentioned, EPRdid notuse this unfortunate conception of
noted, was notsatisfied with stating his metaphysical disagreements
with Einstein—he wanted to “win” the discussion." He therefore de- physicalreality in their paper. Yet their poor wording seemsto call for
a responseto their formulation of the concept of physical reality. From
fined the ground of the discussion as if Einstein and he’shared some
the two notions of reality—the naive realist version of EPR and the
basic presuppositions about how physicists conceive of reality: “The
verificationist version of Bohr—thelatter is the more defensible and
extent to which an unambiguous meaning can be attributed to such an
sophisticated. Byfiat, the reader seems to have no choicebutto follow
expression as ‘physicalreality’ cannot of course be deduced froma priori
the lesser of the twoevils. Einstein appears to be a naive, outmoded
philosophical conceptions, but—as the authors of the article cited them-
scientific realist who “in regard to quantum physics is out to lunch”
selves emphasize—must be founded on a direct appeal to experiments
(Pais 1991, 434). ;
and measurements” (Bohr 1935a, 696, myitalics).
Quite apart from the fact that this poor definition of reality is never
Compare this with the EPR wording: “The elements of the physical
used in the paper, sucha simplistic notion of reality is not Einstein’s.
reality cannot be determined by a priori philosophical considerations,
Einstein himself ridiculed such a naive approach. Einstein wrote: “‘The
but must be found by an appealto results of experiment and measure-
world of bodiesis real.’ ... The above statement appears to meto be, in
ments” (EPR, 777). Note the subtle change in wording and fundamental itself, meaningless,as if one said: ‘The world of bodies is cock-a-doodle-
change in meaning: EPR talked about “elementsof the physical reality,” doo.” It seems to me that‘real’is in itself an empty, meaningless cate-
while Bohr talked about the “meaning”of the notionofreality. In the gory” (quoted in Howard 1990, 368, Einstein’s italics).1* Einstein’s real- -
EPRdiscussion, the elements of physical reality are the physical vari-
ism has been described in Fine’s (1986) pioneering analysis, and the
ables that can be predicted with certainty—one can reformulate the discussion further extended and analyzed in Howard (1990, 1993) and
EPR wording in terms of the adequacyof theoretical notions, without
Beller (forthcoming). For Einstein, the notion of scientific truth is Kan-
invoking the concept ofreality. Bohr reformulated the passage from
tian andholistic: the truth ofa scientific statement does notreside in its
EPRinto a metaphysical discussion of what physicists mean when they
correspondence with reality but derives from the adequacyof the uni-
say “reality.” This reformulation, together with Bohr’s repetition of a
fied conceptual model to which it belongs (the empirical adequacyof
few measurementprocedures, has a strong rhetoricaleffect. Following such a modelis oneof the conditions for its truth; thelogical simplicity
Bohr’s analysis of measurement procedures time and time again, the
of its foundation is another). Einstein's realism avoids the choice be-
readerenters into Bohr’s frame of mind and, without noticing, loses any
tween naive realism and simple-mindedpositivism. -
critical perspective on theverificationist ground that Bohr gradually and
As Fine (1986) has pointed out, the final draft of the EPR paper was’
carefully builds. By tinkering with the wording of EPR, Bohrcreates an written by Boris Podolsky,so it seemsthat it is Podolsky’s rather than
illusion that Einstein, Bohr, and the readerall share the same episte-
Einstein’s conception of reality that is reflected in the opening pages.
mological stand concerning the connection between theory and experi-
Yet, due to the tendentiousrecollections of the Géttingen-Copenhagen
ment. It is on this “common” ground that Bohr “defeats” Einstein.
physicists and to some careless wording in the EPR paper, Einstein’s
The sloppy wordingin the beginning of the EPR paper, where the
conceptionofreality is widely misrepresented and misunderstood. The
authors write about “objective reality, which is independent of any
myth of Einstein the naiverealist and the myth of Bohr’s triumphant
theory,” seems to support Bohr’s ingeniousrhetorical ploy. The opening
defeat of Einstein in their debate over EPR go handin hand.
lines of the EPR paper express a naive, simplistic notion of the “corre-
spondence theory of truth,” where theoretical “concepts are intended
Disturbance, Reality, and Acausality
to correspond with the objective reality” (EPR, 777). Not much philo-
sophical sophistication is needed in order to ask: Whatkindof access
The EPR challenge forced Bohr to make basic changesin his philosophy
do the EPR authors haveto this reality, which is “independent of any
of complementarity, undermining the notion of disturbance on which
theory”? Is not their stand simply a metaphysical prejudice? Is Bohr
his pre-1935 philosophy was built (Fine 1986; Beller and Fine 1994).
notright to combattheir unfounded position? In particular, is Bohr not
Bohr's reaction to EPR was an opportunistic one, in the sense of
right that a more adequate definition of reality is badly needed?
14. On other occasions Einstein also rejected the notion that the “truth” of a theory lay
13. From the point of viewof the second reading, Bohr probablyimaginedinitially in its “correspondencewith reality.” “He[the scientist] will neverbe able to compare his
that picture with real mechanism, and he cannot even imagine the possibility of the meaning
he had wonthe discussion.
of such a comparison” (quoted in Fine 1986, 93).
 1536 Chapter Seven Einstein-Podolsky-Rosen andthe Voices of Bohr’s Response 157

“opportunism”thatI discussed in chapter 3. As I will shortly argue, not reality and acausality problems becamedissociated, and each under-
onlyBohr’s concept of disturbance, but all the basic tenets of Bohr’s phi- went an independent transformation.
losophy—the problemof reality, acausality, and the nature of measure- Bohr’s early description of the nature of measurement invoked the
ment—underwenta complete transformation, even a reversal, from his realistic imagery of existing phenomena(no operational definition of
early (pre-1935) work to his later (post-1935) writings. Bohr’s early and concepts yet!) and of disturbance, or finite changes in the phenomena, .
later philosophical writings cannot be unified into a coherent interpre- during measurement. This idea of disturbanceis elaborated in an early
tive structure. No wonder scholars who assume that Bohr’s philosophi- manuscript by Bohr: “Our usual description of physical phenomena is
cal framework underwent no substantial change over the years (at based entirely on the idea that the phenomena concerned may be ob-
most, some refinement of terminology) experience formidabledifficul- served withoutdisturbing them appreciably. ... the quantum postulate
ties in understanding Bohr’s thought. = ise implies that no observation of atomic phenomenais possible without
After 1935, operationalism becamea focal point of Bohr’s philoso- their essential disturbance” (1927d; BCW,6:91, myitalics).
phy." Bohr (1935b) presented the operational definition of concepts not— This is not an accidental, unhappy choice of terminology—the idea
merelyas plausible but as the only possible one, claiming that the ac- persists in all of Bohr’s writingsin the late 1920s to early 1930s:
count of measuring instrumentsconstitutes the only basis for the defi-
Unavoidable influence on atomic phenomena[is] caused by observing
nition of physical concepts. Despite the fact that Bohrcriticized Heisen-
them. (1929b, 100; BCW, 6:216)
berg’s fallacious reasoningin his description of the y-ray experiment —
in the uncertainty paper, Heisenberg’s-idea of “iuncontrollable dis- Phenomenaare influenced by observation. (1930, 134)
turbance” was the core of Bohr’s pre-1935 writings. It inspired Bohr’s The measurementof the positional coordinate of a particle ig accom-
understanding of the nature of measurement, his stand on the inter- panied notonly byfinite change in the dynamicalvariables. (1928, 103)
relation between observation and definition, and his elaboration The action of the measuring instruments on the object under investi-
ofthe philosophical problemsof causality and reality in the quantum gation cannotbe disregarded. (1933,7) :
domain. Interaction between these instruments and the atom itself [is] an ex-
The conceptof disturbance, inaugurated in Heisenberg’s uncertainty change of such magnitudethatit erasesall trace of the phenomena we
" “paper, is an ill-fated and inconsistent one: it presupposes the existence try to observe. (1935b, 219)
ofobjective exact values that are changed by measurement, contraryto The idea of disturbanceis intimately connected with the immediately
the desired conclusion of indeterminacy. Eventually, Bohr (1939) would appealing —butin fact wrong—ideathat the mere fact that measuring
repudiate the disturbance concept. Bohr’s followers consequently mini- devices are themselves composedof atoms necessitates the inseparabil-
mizedits significancein Bohr’s writings (Pais 1991). Yet disturbance im- ity of atomic phenomenaand the meansof observing them. Bohr often
agery is entrenched in Bohr’s thought at the time. Disturbanceis the emphasized the atomic structure of measuring devices as the reasonfor
reasonfor the inseparability of phenomena and the meansof observing the finitude, or wholeness, of the quantum interaction, which, in turn,
them,for complementarity (rather than alater identity between defini- implies the inseparability of phenomena and observation. The under-
tion and observation), and for Bohr’s initial conflation of the problems lying —incorrect— intuitionis that because in the quantum domain the
of objectivity and causality. According to Bohr, both objectivity and interaction of measuringis of the same order of magnitude as the phe-
causality presuppose the notion of the exact definition of the state of a nomena being measured, such finite interaction (in contrast to the clas-
physical system, excluding in principle all disturbances. Bohr argued sical case) cannotbe neglected, or “accounted for.”
thatif in the quantumdomain every measurement implied intervention, The connection between finite disturbance and the atomic constitu-
or finite, nonnegligible interaction, then the conclusion of inevitabil tion of measuring devices is prominent in Bohr’s pre-1935 writings:
ity
andthe final overthrow of both reality and causality immediately “We cannot close our eyes to the fact that not only the bodies under
fol-
low. Later, when Bohr abandoned the imagery of disturbance, the
16. The idea of disturbance has, of course, a counterpartin the quantum formalism:
15. As Thave argued, in 1927 (the Comolecture) Bohr objected to the operation the change of a quantumstate, or “reduction of a wave packet.” If an atomic system is not
al em-
phasis in Heisenberg’s approach. The definition of concepts, claimed Bohrat the in the eigenstate of an observable, the measurementof this observablealtersits state from
time,is
independentof, and prior to, any procedure of measurement. As I argued
in chapter6, a superposition to the eigenfunction corresponding to the measured value. It is not clear
Bohrheld that the only way to connect the quantum formalism with at all how the rules of the quantum calculus are implied by, or even connected to, the
observable space-
time concepts was through wavetheoretical imagery.
“finitude” of the quantum ofaction.
 158 Chapter Seven
Einstein-Podolsky-Rosen and the Voices of Bohr’s Response 159
investigation, but also the measurement instruments are built
up of between subject and object,” as well as Bohr’s idea of introspection
atoms” (1931b, 152).” The atomic constitution of measur
ing devices im- when no separation between object and subject exists, is inspired
plies not onlydifficulty in separating atomic phenomena and observa by
- his notion of disturbance. Bohrillustrated the “unavoidable”introduc-
tion but a “difficulty of distinguishing between object and
measuring tion of “subjectivity” by describing misleading analogies with relativ-
instruments. With the latter problem weare at once presented whenit
ity "theory: “[The] theoryof relativity ... was destined to reveal the
is necessary to take the atomic constitution of the measuring instru-
subjective character of all the concepts ofclassical physics” (1929b, 97;
ments into account” (1931b, 155). The idea of disturbance implies
a BCW, 6:213).
symmetry between measuring devices and atomic objects: “The
idea After 1935, asa response to the EPR challenge, Bohr abandoned the
of the means of observation independentof the phenomenaor of phe-
terminology and underlying imagery of disturbance. The problems of
nomena independentof meansof observation cannot be maintained”
reality and causality became dissociated, and Bohr made an about-face
(1927d, 91)—a claim that stands in striking contrast to Bohr’s
later in his opinions on the nature of measurement. In contrastto the earlier
Position that the nature of atomic objects and the nature measur
ing de- symmetry between the measuring and the measured, Bohr introduced
vices are fundamentallydifferent. ; the idea of a fundamentaldistinction between the nature of atomic ob-
The idea of disturbance, of “interference” with the course of phe- jects and that of measuring devices. In Bohr’s later writings an experi-
nomena, underlay Bohr’s pre-1935 writings on both objectivity and cau-
mental device must beclassical, “heavy,” and its atomic constitution
sality. The discoveryof the quantum ofaction throws “a new light upon mustbe, “in principle,” disregarded. The measurementinteraction can-.
the old philosophical problem of the objective existence of phenomena not be separated from phenomena,not because one cannot neglect the
independently of our observations. Any observation necessitates an in- quantum, but precisely because one must neglectit: “The essentially
terference with the course of the phenomena. . . . the limit of speaking about new feature in the analysis of quantum phenomenais, however, the
phenomenaasexisting objectively finds its expression . . . just in the introduction of the measuring apparatus and the objects under investiga-
formulas of quantum mechanics” (1929a, 115; BCW, 6: 249; myitalic
s). tion. This is a direct consequence of the necessity of accounting for the
Bohr'’s understanding of the breakdown of objectivityinitially implied
functions of the measuring instruments in purely classical terms, ex-
the introduction of a perceiving subject into physics ’*—an idea that
cluding, in principle, any regard-to the quantumof action” (1958c, 3,
Bohr would later deny vigorously (when he identified the act of mea- Bohr’s italics). The crux of Bohr’s later argumentsis precisely that the
surement with the permanenceofthe recordings of measuring devices). measurement interaction is nonformalizable in principle.” .
The “close analogy” Bohr drew between the impossibility of strictly :
As Bohr’s ideas on the nature of measurement took an about-face,his
separating phenomenaand observation and the “general limits of man’s accounts of objectivity and causality changed fundamentally. In Bohr’s
capacity to create concepts which have their roots in our differentiation early writings the main psychological analogue of the inseparability of
17. The atomic structure of measuring instruments, according to Bohr, phenomenaand observation wasintrospection, in which no sharp dis-
does not merely
havefar-reaching epistemological consequences; a more thorough and explicit tinction can be made between object and subject. Asin psychology, so
incorpo-
rationofthis fact will eventuallylead to further developmentof the
physical theory itself: in physics, the idea of objectivity, which avoids any referenceto a per-
“Thepresent formulation of quantum mechanics in spite of its great fruitfulness ceiving subject, can no longer be maintained. Bohr’s implicit definition
would
yet seem to be no morethan first step in the necessary generalization of
mode of description, justified only by the possibility of disregarding in its
the classical of objectivity wasinitially metaphysical—the existence of objects and
domain of events having an independentreality regardless of being observed or
application the atomic structure of the measuring instruments themselvesin
the interpre-
tation of the results of experiment” (1937c, 247). A similar idea was develope not. In contrast, Bohr’'s later definition of objectivity was intersubjective:
d in Bohr’s
(1937d) Hitchcock lectures at Berkeley. Bohr identified an objective description with a consistent methodof re-
18. “The feature which characterizes the so-called exact sciencesis, counting facts that can be understood clearly by others. After 1935,
in general, the at-
temptto attain to uniqueness by avoidingall reference to [a] perceivin
endeavoris found most consciously, perhaps,in mathematical symbolism
g subject. This Bohr was eager to retain objectivity, not to dispense withit. It is not
which sets up clear how the use of imprecise common language ensures unambiguity
for our contemplation an ideal of objectivity to the attainment of which
scarcely anylimits
are set, so long as we remain within a self-contained field of appliedlogic.
In the natural 19. “It is decisive to realize . . . that the description of the experimental -arrange-
sciences there can be no question of a strictly self-contained field of applicatio
n of the ments . . . must be based on commonlanguage. . . . This circumstance . . . excludes any
logical principles, since we must continually count on the appearanc
eof new facts” (Bohr separate accountof the interaction between the measuring device and the atomic objects
1929b, 96-97, BCW, 6:212-13).
underinvestigation” (Bohr 1960c, 773).
 160 Chapter Seven
Einstein-Podolsky-Rosenand the Voices of Bohr’s Response 161

of communication—one can easily argue that esoteric technical lan- extension of our intuition is, however, both historically and philosophi-
guage is more suitable, as Heisenberg,in fact, did (see chapter8).
cally inaccurate. This claim disregards the tremendousgap between our
This change in the conception of objectivity, from observer inde-
essentially Aristotelian intuition and the sophisticated, abstract frame-
pendence to unambiguous communicability (a change forced by EPR), work of Newtonian physics (and thus ignores the huge intellectual
correspondeddirectly to Bohr’s about-face on the nature of measure- effofts of the founders of modern science, who replaced the intuitive
ment: communication is only unambiguousif we put a separation line
Aristotelian world with the counterintuitive Galilean-Newtonian one).
between subject and object. Consequently, Bohr’s analogies between Noris it obvious that the mostessential feature of a measuring device
quantum physics and psychology change in character, from the ex-
is the classically described space-time configurationofits components.
ample of introspection,in which noseparation between subject and ob- Dirac, for example,atleastinitially, was at odds with Bohr onthis point.
ject exists, to the famous case of Meller’s student, who becomes con- |
As Oskar Klein recalled: “Who did not quite agree with Bohr atfirst
fused trying to untangle his different selves. Some students of Bohr’s was Dirac. . . . he made somekind of observation philosophy which had
thought read this exampleincorrectly; Bohr advancesit to argue.that to do with permanent marks. Bohr objected lot to that becauseit . . . is
the line between object and subject must be drawn sharply, not that the
not the essential thing in the observation. But Dirac madethat, so to say,
two are inseparable.If we blurtheline, as the poor philosophystudent as the essential thing . . . because one part of Bohr’s view was that you
does, we becomeincapacitated, unable to function in dailylife and to had to... use the whole classical theoryin describing observation. And
look for employment. Eventually, we may evenfall into madness: “This ' [ think that was not near to Dirac’s mind at the time” (interview with
situation can give-rise to what we.call splitting of personality.” The Klein, AHOP). ae
chilling moral of Meller’s humorousstory is “how essentialit is to pay The historical context of Bohr’sinitial idea of the indisperisability of
attention to the separationlines as for example, in physics, separation classical concepts wasas follows: As I have argued, the biggest philo-
between system and the observer” (1958b, 715). sophical quandary of the new matrix mechanics involved its elimina-
For the later Bohr, separating the “content of our consciousness” and
tion of the space-time container from the atomic world (chapter 2).
“the backgroundloosely referred to as ourselves” is not only not im- Heisenberg, Born, and Jordan accepted lack of visualization as the
possible—it is mandatory for “unambiguous communication,” even necessary price—and perhaps, in fact, a bonus—of an outstanding
- though the separation lines may be placed differently in various con- technical advance. Even Bohr became “more and miore convinced of the
texts due to the “richnessof the reality of consciouslife” (1960c, 13). need of a symbolization if one wants to express thelatest results of
physics” (Bohr to Harald Hoffding, quoted in Jammer 1966, 347).
Bohr’s Doctrineof the Indispensability of Classical Concepts The great success of Schrédinger’s competing version of quantum
and the CorrespondencePrinciple mechanics changed Heisenberg’s stand on this issue. As Klein recalled,
throughoutfall 1926 and winter 1927, Bohr and Heisenberg searched
Bohr’s doctrine of the indispensability of classical concepts underlies feverishly for a way “to introduce space and time into these complex
his philosophy of complementarity. Give up this doctrine, and the in- formulae” (interview with Klein, AHQP). Heisenbergfinally came out
evitability of complementarity in physics dissolves. Even today, the with the answerin his uncertainty paper—allclassical space-time con-
most sympathetic interpreters of Bohr’s thoughtdo notfeel at ease with cepts can be retained in the quantum domain if one gives up their pre-
Bohr's categorical assertions of the impossibility of concepts other than cise simultaneous use; yet, in contrast to classical theory, variables in
the classical (Hooker 1994; Howard 1994). In the past, some of Bohr’s
the quantum domain are subject to the uncertainty relations.
closest collaborators, Heisenberg and Born, rejected this peculiar doc-
What for Heisenberg wasan exercise in possibility (though at times
trine (see chapter8).
he confused it with the rhetoric of inevitability; chapters 5 and 9) be-
Bohr presented his doctrine of the indispensability of classical con-
came, for Bohr, the argumentfor indispensability. The indispensability
cepts in a deceptively convincing way. Classical concepts, Bohr ar-
of classical concepts was discussed thoroughly in the correspondence
gued, being direct extensions of our intuition, are necessary to describe
between Bohr and Schrédinger at the time. Bohr’s dismissal of the pos-
experiments and to communicate to others what we have done and
sibility of developing “new” concepts was expressed initially as an
learned. Bohr’s claim that classical concepts constitute a necessary
argument against Schrédinger’s plan to search for a new conceptual
 162 Chapter Seven
Einstein-Podolsky-Rosen and the Voices of Bohr’s Response 163
theoretical scheme that would avoid quantum theory’s peculia
rities and In striking contrast to his later analyses, before 1935 Bohr claimed
“irrationalities.’’?°
When Bohr wrote thatthe “interpretation of the experimental that only by incorporating the atomic nature of measuring devices more
mate- thoroughly into the analysis of the microdomain could the interpreta-
rial rests extensively upontheclassical ideas” (1927d,
91) or that“it lies tive program be advanced. After 1935 Bohr moved graduallyto an in-
in the nature of physical observationthatall experience mustul
timately termediate position, according to which one can “to a very high degree of
be expressed in termsofclassical concepts, neglecting
the quantum of approximation disregard the molecular constitution of the measuringin-
action” (1929b, 94-95; BCW, 6: 210-11), he was
not referring to the idea
of treating measuring devices as classical in nature, struments” (1948, 451, myitalics). This “approximation” later became
as some commen-
tators have understoodit, reading Bohr’s later ideas backwa a matter of principle—it was nowa “logical” necessity to ignore the
rd. Instead, atomic structure of measuring devices and describe them in “purely
Bohr was making a broad Kantian statement about the impossi
bility of classical terms, excluding in principle any regard to the quantum of
describing physical experience in general by any concepts
other than action” (1958c, 3-4). This “essentially new feature” in the analysis of
classical ones.
As in Heisenberg’s uncertainty paper, so in Bohr’s quantum phenomenais tied directly to the “heaviness” of the measur-
numerous discus- ing devices as opposedto the microobjects: apparatus are “sufficiently
sions of thought experiments in his pre-1935 writings, classical con-
cepts apply simultaneously to microscopic objects and macros large and heavyto allowan accountof their shape and relative posi-
copic tions and displacements without any regard to any. quantum features
measuring devices. In Bohr’s early writings, he spokeof the “preestab-
lished harmony” between quantum theoretical concepts and the pos- inherently involvedin their atomic constitutions” (1962, 24). -
sibilities of measurement: a measuring device seems to be a perfect Similar passagesare scattered throughout Bohr’s later writings (Bohr
match to the wavetheoretical definition of particles: “Uncertainty 1958b, 712; 1958c, 2—4). This “heaviness” is now crucial for unambigu-
can- ous communication, and, indeed, in orderto talk about well-defined
not be avoided. .. . the experimental devices . .. are seen to
permit only experimental conditions at all.” After 1935 Bohr held that the “heavi-
conclusions regarding the space-time extension ofthe . . wavefields
[associated with particles]” (1928, 101)” ness” of the measuring apparatus, andthe fundamental distinction be-
tween measuring and measured that follows from it, is the reason phe-
The measurementinteraction in Bohr’s early writings was quant
um nomena and observation are inseparable (1958b, 712, myitalics).”
mechanical— quantum uncertainty applied even to the whole macro-
scopic y-ray microscope as a measuring device: “A closer investigation This fundamental distinction between the measuring and the mea-
shows, however, that such a measurement [of the momen sured undermined Bohr’s earlier arguments about the complementarity
tum transmit- of space-time and causality. These arguments relied on the “uncon-
ted during measurement] is impossible, if at the same time one
wants trollability” of the measuring interaction, which in turn was demon-
to know the position of the microscope with sufficient accuracy.In
fact, strated by applying the uncertainty relations to the measuring devices
itfollows from the experiences which havefound expression in the
wave theory regarded as objects of measurement.” But, once Bohr introduced the
of matter that the position of the center ofgravity of a body and
its total mo-
mentum can only be defined within the limits of reciprocal accura
cy given by 22. “In the description of the experimental arrangements we must certainly use our
relation (2) [the uncertainty relation]” (my italics). Bohr summar
ized: ordinary language. . . . We can by experiments only understand something about which
“The uncertainty equally affects the description of the weare ableto tell others what we've done and what we've learnt. . .. The reason that one
agency of mea-
surement andofthe object’’ (1928, 101, 102). can describe experimental conditions in this matter is that one uses as apparatus heavy
bodies, bodies which are so heavy, immensely heavy, compared with the single atomic
20. “Tam scarcely in complete agreement with yourstress particles that we can, in the description entirely neglect all implications of the quantum”
on the necessity of devel- (Bohr 1958d, 727).
oping ‘new’ concepts. Not only, as far as I can see, we have up
to now no cues for sucha
rearrangement, but the ‘old’ experiential concepts seem to 23. It is this unbridgeable gap betweentheclassical and the quantum that implies the
me to be inseparably con-
nected with the foundation of man’s power of visualizing” unsurveyability, or unformalizability, of the interaction, implying in tum a “wholeness,”
(Bohr to Schrédinger, 23 May
1927, AHQP; quoted from Murdoch 1987, 101). or the necessity of specifying the experimental conditions for any definition of quantum
21. A similar idea is expressed in the following passage: phenomena (Bohr 1961, 77-78). This gap is made even larger as Bohr talks of the neces-
“Limitation on the possibili-
tiesof measurementis directly related to apparentcontradictions sity of using not even “classical” but only “plain” or “ordinary” language (1958c, 2-3:
in the discussion ofthe
nature oflight and of material particles” (contradictions that 1958e, 695). .
are removed,according to
Bohr, by using the wavetheoretical definition of particles; 24. Thus, in EPR, one cannot use the exchange of momentum between a suspended
Bohr 1929b, 95; BCW,6:211).
diaphragm anda particle to predict the exact value ofthe particle’s momentum,because,
 164. Chapter Seven
Einstein-Podolsky-Rosen and the Voices of Bohr’s Response 165
idea that the fundamental distinction between the measuring and
the measured is tied directly to physical constitution (“heaviness”), the discovery of the new quantum theory, mathematical physicists (Som-
same object simply cannot be used asa measuring device and as a mea- merfeld, for example) perceived correspondence arguments as a tem-
sured object even in different contexts.” In this case, not only Bohr’s porary heuristic tool, criticizing Bohr’s elevation of the principle into
arguments about the complementarity of space-time and causality but a ae of the quantum theory” (Mehra and Rechenberg 1982). After
the whole early enterprise of demonstrating compatibility between the discovery of matrix mechanics the natural conclusion was that the
the consistent mathematical formalism should supersede Bohr’s correspon-
quantitative conclusions of the quantum formalism (uncertainty ‘rela-
tions) and the experimentalsituation (analysis of measurements) is un- denceprinciple. Yet Bohr never abandonedit, both as a heuristic and as
dermined.It is only in some vague sense that one can claim, as Bohrdi an interpretive principle. Bohr had no doubtthat correspondence con-
d - siderations must guide the developmentof quantum electrodynamics.”
after 1935, that the unsurveyability of the measurementinteraction and
the consequent “wholeness” of quantum phenomenadictate Andit served as a touchstone in his later writings: “It is of course
a depar- important in each case to remember howfar from the point of view
ture from classical description (see chapter9). For example, itisnot clear
in Bohr’s later writings why this “wholeness” indicates the necessityof of correspondence we can acquaint ourselves with the new situation”
statistical description—in fact, it is not clear howone assesses the nec- (1957c, 666). a
essaryfeatures of an appropriate descriptionatall (see chapter 12). Bohr . Bohr's insistence on the continuing importanceofthe correspondence
simply retained an intuitive connection between the necessity of clas- principle seems puzzling. Dirac, who had himselfinitially used corre-
sical concepts, the inseparability of phenomena and observation, and - spondence arguments in a most ingenious way (Dartigol 1992a), did
not share Bohr’s attitude: “He [Bohr] still referred to the correspon-
acausality—ignoring their separate sources in now incompatible argu: —
ments. This maneuver has resulted in the misconception that Bohr’s denceprinciple for some years, I think, after quantum mechanics really
philosophical framework remained unchanged. madedefinite equations which would replace the correspondenceprin-
While the meaning of Bohr’s doctrine of the indispensability of clas- ciple. ... When one gets so absorbed with one idea, one does stick to it
sical concepts varied over the years, there is a sense in which it re- always” (interview with Dirac, 14 May 1963, AHQP). Why did Bohr
mained unchanged. The claim that classical concepts are necessary is need correspondence arguments to acquaint himself “with the new
-- connected with Bohr’s principle of correspondence—a principle that situation”? Whydid he notrefer in his writings to featuresof the quan-
guided Bohr’s early work in quantum theory, and to which Bohr sub- tum formalism, and to such associated interpretive schemes as the pro-
scribedall his life. Bohrhad introduced the correspondence principle jection postulate? Whydid he never consider the quantum formalism
as a heuristic guide, as a methodologicalprinciple of research—a “for- with articulated bridge principles a viable interpretive option?
mal analogy between the quantum theory andtheclassical theory.” The I can suggest part of the answer: For Bohr, the necessity of the corre-
correspondence principle stated, not only that for large quantum num- spondenceprinciple and the indispensability of classical concepts were
bers classical and quantum calculations should coincide, but also that not merely the subjects of an abstract philosophicalinquiry. The purely
for small quantum numbersthere is a correspondence between various philosophical considerations were invoked to “dress up” a method-
harmonic components of motion computed classicaily and the charac- ological question— the questionof the success ofa particular program
teristics of various types of quantum transition from one stationa of investigation. His heuristics, which he internalized because of long
ry and successful use by himself and others, eventually became the most
state to another (Jammer 1966; Hendry 1984; Darrigol 1992a).
Bohr'’s correspondence principle, applied with ever increasing rigor essential feature of all possible research programs for Bohr.”It is
by Kramers, Born, and finally, in a fundamental way,
by Heisenberg, 26. “That for the moment the paradoxes connected with the use ofthe idealization of
led to the formulation of the new quantum formalism. Even point charge for the electron are preventing the development on correspondencelines of
before the
a comprehensiverelativistic quantum electrodynamics, must indeed rather be imputed
in order to measure a diaphragm’s momentum,“this body can to our failure .. . to grasp some deeper feature of the stability of the individual particles
no longer be used as a
measuring instrument ... but must. . . be treated,like the particle themselves than to any lack of soundnessof the generallines on which the incorporation
traversing theslit, as
an object of investigation, in the sense that the quantum of the quantum of action in atomic theory has been achieved”(Bohr 1939, 389).
-mechanical uncertainty rela-
tions regarding its position and momentum must be explicitly taken 27. In this sense, Bohr is similar to Einstein. An enlightening analysis of Einstein's
into account’ (Bohr
1935a, 698). : philosophical makeup was given by Arthur Fine (1986), who argued that for Einstein
25. This fundamental distinction is not merely
a semantic one (the “cut” as discussed
philosophical questions about “realism” and “determinism” were not abstract philo-
in chapter9) of the context in which, for example, the diaphra
gm is used. sophical questions but questions of the success of research programs assuming these
notions.
eoaes
 166 Chapter Seven
Einstein-Podolsky-Rosen and the Voices of Bohr’s Response 167
precisely because Bohr did not (and possibly could not—see
chapter cannot determine the choice between different individual processes”
12) participate in the mathematical elaboration and consoli
dation of (1948, 446). In an earlier paper Bohr argued that “quantum postu-
quantum theory that classical theories with their underlying
mathe- Jates ... imply an explicit renunciation of any causal description. ...as
matical structures appeared to him the only tools one could useto
grap- regards its possible transitions from a given stationary state to another
ple with the quantum world. In this way, personal idiosyn
crasy was stationary state . . . the atom may be said to be confronted with a choice for
transformed into an overarching principle of knowledge. Since
the which, according to the whole character of the description, there is no
claim that classical concepts are necessary is the least defensi
ble of determining circumstance” (1939, 385, myitalics). Yet, in the same pa-
Bohr’s assertions,its rhetorical underpinnings are especially strong.I
t per, a few pages later, Bohr used Heisenberg’s argument for acausal-
is usually presented as “obvious,”“as a “simple logical demand”(1948;
ity, incompatible with the position just cited: “The essentially statistical
1950, 512-13; 1958d, 727). . nature of this account[is] a direct consequence of the fact that the com-
[ have arguedthat it is impossible to combine Bohr’s pre-1935 and
mutation rules preventus to identify at any instant more than a half of
post-1935 writings into a unified, coherent structure, and that a basic
change occurred in Bohr’s notions of disturbance, reality, acausali
the symbols representing the canonical variables with definite values of
ty, the correspondingclassical quantities” (1939, 387).
and the indispensabilityof classical concepts. It is tempting to assume
I will argue in the next chapter that similar changes and contradic-
that perhaps Bohr’s early andlater writings, taken separately, are sys-
tons characterize the interpretive pronouncements of other quantum
tematic and free of contradictions. Yet textual evidence does not sup-
physicists. The Copenhageninterpretation waserected, not as.a consis-
port this assumption either. Bohr’s discussions of causality are a case
in tent philosophical framework, but as a coliection of local responses to
point. : changing challenges from the opposition.
Despite Bohr’slifelong preoccupationwith the issue of causality, his
use of this concept was unsystematic and contradictory. His, notion of
causality was very “thick”: Sometimesit was cause-effect relationship.
Sometimesit was “determinism.” Sometimes it was an epistemological,
other times an ontological, definition. Sometimes causality was equated
with the applicability of the conservation laws of energy and momen-
tum, other times with the simultaneous applicability of space-time and
energy-momentum concepts. Sometimes Bohr’s understanding of cau-
sality was probabilistic, applied to an individual system; and some-
times it wasa statistical interpretation, applied to an ensemble of simi-
lar systems. Bohroften conflated determinism and predictability, which
are in fact different notions.
In Bohr’slater writings, he often employed a definition of causal de-
scription as predictability: “In physics, causal description . . . rests on
the assumption that the knowledgeofthestate of a material system at
a given time permitsthe predictionofits state at any subsequent
time”
(1948, 445). Note that such a definition of “causality,” and
correspond-
ingly of “acausality,” is compatible both with the idea that the lawsof
nature themselvesarestatistical (Born’s view) and with theposition
that
the laws of nature are deterministic andstatistics is introduced only
because of an inability to determine simultaneously and exactly all the
physical variables of a state (Heisenberg’s view). Bohr employed
both
of these mutually contradictory concepts of acausality at roughl
y the
same time. He spoke in an ontological vein, about a “free
choice of
nature”: “The specification of the state of a physical system
evidently
 PART TWO
Sy

Rhetorical Consolidation
 CHAPTER 8
SS
The Polyphonyof the
Copenhagen Interpretation and
the Rhetoric of Antirealism. ..

All that mattered now was to present the facts in such a way.that, despixe
their novelty, they could be grasped and acceptedbyall physicists.
Werner Heisenberg 1971, 79

Introduction

If intellectual activity is by nature addressive and communicative, if

thoughts are suffused with the voices of visible and invisible inter-
locutors, a dialogical analysis is bound to reveal polyphony and in-
coherence in ascientist’s philosophical accounts. Solid structures and
conceptual frameworks, presumably built on stable, paradigmatic pre-
suppositions, turn out to be ahistorical. abstractions or misleading il-
lusions.
_ The Copenhagen paradigm is sepposediy built on two incontestable
pillars—indeterminism anda revision ofthe classical notion ofreality.
Wehave seen that the conceptof indeterminism hadnofixed meaning
for the Géttingen-Copenhagen physicists: Bohr, Heisenberg, and Born:
each used different notions of causality in different theoretical and
social circumstances. If no coherence can be found among all the ut-
terances on a “given issue” made by a single person,it is unreason-
able to expect that the writings of a group of thinkers will cohere into
a unified consistent framework. I have argued that Bohr’s notion of
complementarity in his Como lecture was not compatible with Heisen-
berg’s particle ontology. Similarly, wave-particle complementarity was
not compatible with Born’s probabilistic interpretation of quantum me-
chanics.
Certain positions seem to be morestable, or perhaps more recurrent,
than others—Bohr’s wave-particle complementarity, which Heisenberg
and Born often supported publicly, is a case in point. Yet such positions
only appear to be stable because they are repeatedly presented to lay
 172 ChapterEight
The Copenhagen Interpretation and the Rhetoric of Antirealism 173
audiences, from whoma serious scientific challenge
can hardly be ex- ence.”? Thefirst mythology assumesthat philosophical writings neces-
pected. The variety of audiences to which orthodox quant
: um physi-
cists addressed their statements was a major source of contra
dictory
sarily deal with “eternal” epistemological or ontologicalissues, such as
ie
t
elements in their writings. Heisenberg adopted Bohr’s positiv
ist ap-
the “realism” of atoms. Bohr, for example,is often classified as a “real-
ist” because he “never doubts” the reality of atoms. The second my-
proach for mathematically unsophisticated audiences, yet he
employed tholdgy assumesthat philosophical writings are in principle intended
elements of a realistic ontological interpretation when
addressing his to produce a systematic contributian to their subject, and that to “un-
mathematically skilled colleagues. Sometimes,as I argued
in chapter 5, derstand” such writings means to produce, at whatever cost, the most
two contradictory approaches were present simultaneously
: Contra- coherent and contradiction-free presentation of their content.I suggest
dictions and inconsistencies in the writings of the foundersof the quan-
tam.revolution are not an anomaly, but the order of the that such “mythologies” are especially inapplicable to interpretive ex-
day. In this plorations by working scientists, who espouse their philosophical po-
chapter I will articulate this point with respect to two centralpillar
s of sitions in dialogical responses occurring in changing theoretical and
the orthodox interpretation: antirealism and the-necessity of
classical sociopolitical circumstances. I argue that different pronouncements by
concepts. 2 8 a ge : quantum physicists on the realism-antirealism issue bai be understood
Strong realistic and positivist strands are present in the writin
gs of by differentiating between what scientists “need not and mustnot
the founders of the quantumrevolution—Bohr, Heisenberg, Pauli,
and do. “Need not” is a notion linked to addressivity, which is aptto head
Born. Militant positivist declarations are often followed by
fervent de- off potential and existing objections to a scientific innovation;“must
" nials of adherence.to positivism. No. wonder different scholar
s, with not” is an overarching binding dictum that physicistsabstractillegiti-
good textual evidence, have arrived at conflicting interpr
etations of mately from thelocal “need not.”It is the local “need not,”rather than
‘these writings. While Popper (1963b) presented Bohr as a “subjectivist,
” the universal “must not,” that turns out to be most relevant to under-
Feyerabend (1981a) foundhim an “objectivist.” More recently, Mur- standing the philosophical deliberations of scientists. o
doch (1987) has concluded that Bohr was a realist, while Faye
(1991) has I also argue that Bohr’s antirealism served as a potent cohesive force
‘arguedwith equal competence that Bohr wasan antirealist. Whethe
r inside the community of physicists, and as a way to prevent its alien-
Bohr's philosophy was“realist” or “antirealist” is a topic of great inter-
ation from widercultural circles. For this reason,I suggest, Heisenberg
est, as is demonstrated by a recentcollection of essays (Faye and Folse
and Born often supported Bohr’s position, one incompatible withtheir
1994). Although scholars have invested much serious thought
and in- own. This public espousal of antirealism is one of the major sources of
genuity in rendering Bohr’s position consistent, I hold that conflic
ting contradictions in Heisenberg’s and Born’s writings. Under closer scru-
opinions of realism and positivism (in Bohr’s instrumentalis
t or Heisen. tiny the Copenhagen paradigm has neither coherence nor stability, de-
berg’s operationalist version) are both undeniably present
. My aim is spite the massof rhetoric by Bohr andhis followersassertingits ‘inevi-
not to cure this “schizophrenia”(the characterization in Fine
1986) by tability.”? The Copenhagen philosophycan thus be seenas a contingent
eliminating the inconsistencies but to analyze the sources
, uses, and composite of different philosophical strands, the public face ona hid-
aims of such shifting philosophical positions, on both the concep
tual den web of constantly shifting differences amongits founders.If philo-
and the sociological level.
sophical pronouncements by quantum physicists are most adequately
This: chapter further elaborates the stand, now commonplace
in the understoodas local, shifting, and opportunistic, the question arises of
historiographyof science, that ideas can properly be unders
tood only how the appearance of consensusis achieved. The concluding pages of
by an analysis of their local theoretical and sociopolitical
and use.’ It also supports an approachcritical of what
emergence this chapter addressthat issue.
Skinner (1969)
christ ened a “mythology of doctrines” and a “mytho
logy of coher- WhatScientists “Need Not” and “Must Not” Do
1. This stand, from Collingwood (1939) and Wittgenstein
(1953), through Rorty (1979) Oneof the most useful hints to how to deal with the realism-positivism
and Fine (1986), now has an overwhelming following
among scholars dealing with social
studies of science and with the rhetoric of science.
Some prominent examples include
problem can be found in Feynman’s distinction between whatscientists
Collins (1992), Shapin
and Schaffer (1985), Knorr-Cetina (1981), Galison
(1987, 1997), La- 2. Tam grateful to Menachem Fisch, who made me awareof the similarity between the
tour (1987), and Pickering (1992). However,
these intellectual currents hardly touch the
historiographyand philosophy of quantum mechanics. Oneofthe ideas expressedin thefirst draft of this chapter and the writings of Skinner.
early quantum physics —Darrigol( 1992a)— exemplif best recenthistories of 3. For an analysis of Bohr’s rhetoric of inevitability, see Beller (1993, 1996a) and chap-
ies this point. ter 9.
 174 Chapter Eight
The CopenhagenInterpretation and the Rhetoric of Antirealism 175
“must not” and “need not” do (Feynman, Leighton, and
Sands 1969). It principle. Theoretical physicists “need not” seek such a description be-
is wrong to say that scientists “must not” presupposere
ality, aiming at cause Born’s own probabilistic theory is sufficient, being in perfect ac-
the elimination of all unobservable theoretical entities. Feynm
an stated cord with the practice of experimentalists: “Earlier, it was supposed . .
that no creative scientific theorizing can proceed in such an imagin
ative that there is meaning in the question: ‘When and wherean election is
vacuum. Yet, when certain realistic and seemingly indisp
ensable no- ejected’ ... Suppose that we decide to renounce this question, an act
tions lead to contradictions, one can eliminate
them, provided a consis- whichis the easier inasmuch as no experimenter would think of asking
tent interpretation of experiments is possible without their
use. In the it” (Born 1928, 32).¢ ws
two-sli t experiment, one “need not” assume that a particl
e traverses In order to protect quantum theory from the charge of incomplete-
a well-defined path between the.two-slit diaphragm and the
detector ness, or perhapsof inconsistency, Bohr counteredtheEPR challenge by - -
(Feynman, Leighton, and Sands 1969, 38-8, 38-9). “Need
not” is an in- arguing that one “need not” assume that certain atomic attibutes exist
tegral part of scientific practice, without which such breakt
hroughs as independent of the process of measurement (chapter 7). Reality be-
the rejection of absolute simultaneity in relativity theory and
the rejec- comes, by definition, “what quantum mechanicsis capable of describ-
tion of a strict deterministic framework in quantum
theory would not ing” (Rosen’s remark, quoted in Heilbron 1987, 219). Pauli and Heisen-
have been possible. “Must not”is a positivist excess, at odds
with the berg (who often supported Bohr publicly) did not feel at ease with an
practice of science, which relies on realistic inspirations
as a source of explicit positivist approach. We should. not confuse their-reluctant con-
motivation for research, and onrealistic models as a heuristic guide
to cession to a positivist legitimation with the voluntary embrace of a posi-_
discovery.‘ . : ,
tivist stand. —
Much confusion on the realism-antirealism issue in quantum phys-
ics is caused by the failure to discriminate between the local “need According to Pauli, one does not have to be an empiricist in order to
not” and the overarching “must not.’ When Heisenberg dispen realize that certain realistic forms of thought (Denkformen) are not nec-
sed essary for the successful construction of a physical theory (1954, 123).
with electron orbits and the space-time container in his reinterpreta
tion Heisenberg’s notion of “practical” versus “dogmatic” realism was simi-
paper, he did so to obtain a novel solution, incompatible with the
ex- larly aimed to provide enough leeway for a positivist protection of
istence of such orbits. He realized that he “need not” keep the
elec- quantum theory from. possible objections, without thereby losing the
tron orbits, because they “cannot be observed anyway”(Heis
enberg to realistic ground altogether (1958, 81-83). According to Heisenberg,
Kronig, 5 June 1925, AHQP). As Ihave argued, Heisenberg
adopted the practical realism and dogmatic realism differ in the way they objectify
principle of elimination of unobservables a posteriori, in order
to legiti- experience (the English translation coins the word “objectivate’”). A
mate his unconventional, formal solution. Yet Heisenberg soon
realized statementis objectified if its content does not depend on the conditions
that he had beentoo hasty in elevating the local “need not” to
the sta- of its verification. Dogmatic realism assumes that there are no state-
tus of universal “must not” and changed his stand accordingly (chap-
ters 2, 3, and 4).° Born’s positivist statements are also best
ments concerning the material world that cannotbeobjectified. Practi-
understood cal realism assumes that many,butnotall, statements can be objectified.
on thelevel of the local “need not.” When Born provided a probab

Bet ay
i- Practical realism is indispensable for scientific work: “Every scientist
listic solution of the collision problem anda probabilistic interpretati
on who does research work feels that he is looking for something thatis
for the wave function, he did not hold that a more detailed determ
in- objectively true.” Practical realism, according to Heisenberg, “has al-
istic description in terms of microscopic coordinates
is forbidden in ways been and will always be anessential part of natural science.”
What quantum mechanics demonstratedis that good science is possible |
4. On “motivational realism,” see Fine 1986. MacKin
non (1977) arguedthat the virtual without a basis of dogmatic realism (Heisenberg 1958, 82). While Hei-
oscillator heuristic model was crucial in determining
Heisenberg’s road to the new quan-
tum theory. On the strong presenceof realist notions in senberg’s pragmatic notion of practical realism is not compatible with
what have usually been taken Bohr’s operational definition of physical phenomena,both are directed
as developments characterized by positivist and instrume
ntalist philosophy, see Beller
(1992b). toward the same aim—to unburdenphysics from those troublesome
5. Heisenberg admitted in his Chicago lectures: “To avoid
these contradictions [be- realist notions that “need not” be presupposed in quantum description
tween quantum theory and experiment], it seems necessa
ry to demandt hat no concept and to ease the acceptance of a new and controversial knowledge.
enter a theory which has not been experimentally verified
. . . . Unfortunately, it is quite
impossibleto fulfill this requirement” (1930, 1-2).
6. Page references for Born’s papersare to the collection Born (1956).
 176 Chapter Eight
The CopenhagenInterpretation and the Rhetoric of Antirealism 177
The Appeal of Antirealism: Some General Considerations
scientific community. The discord that can result from competition be-
Antirealist philosophical stands have wide sources, both philoso tween different models, each claiming a privileged roadto truth, was
phi- eloquently displayedin the fierce and bitter confrontation between Hei-
cally andhistorically. The appeal of antirealism varies substantially
in senkerg and Schrédinger over the superiority of their equivalenttheo-
different theoretical and sociopolitical contexts. In the
past, antirealist ries. Bohr’s Comolecture, and the subsequentantirealist philosophy of
sentiments were often invoked because it had been recognized
that complementarity, was intended not onlyto “harmonize” the discordant
equivalent accounts of the same domain of phenomenawere possibl
e. theoretical notions but to pacify the antagonized factions in the com-
Demonstrations of the possibility of equivalentdescriptions,
by Hip- munity of physicists: “I shall try . . . to describe to you a certain general
parchus inantiquity, Poincaré in the nineteenthcentury, Schréd
inger point of view ... which I hope will be helpfulin order to harmonize the
and Pauli in the twentieth century, strengthened the notion of science
as merely a device “to save the phenomena.”” Thus P. W. Bridgm apparently conflicting views taken by differentscientists” (1928, 87).
an Pauli, the (sometimes) hidden architect behind the major interpretative
argued that the “quest for underlying reality” is a meaningless
one, breakthroughs in quantum physics, understoodthe crucial importance
whenonerealizes that two explanationsof the same phenomena
can be of Bohr’s unifying mission. “Niels Bohr,” he announced, “integrated, in
equally true: “However much one might have been inclined fifty years
ago to see some warrant for ascribing physical reality to the lectures and international congresses, and at those carefully planned
internal conferences in Copenhagen,the diverse scientific standpoints and epis-
processes ofa theory .. . certainly no one of the present genera
tion will temological attitudes, and thereby imparted [to the physicists] . . . the
be capableof so naive an attitude after our illuminating experi
ence of feeling of belonging, in spite of all their dissensions, to onelarge fam-
the physical equivalence of the matrix calculus and the wave mechan
- ily” (1945, 97). ,
ics” (1930, 21). er 8
Theattraction of antirealist attitudes becomes strongwhenoneis ea- As the headof this large family, Bohr did not tolerate quarrels. He
ger to avoid a clash between thefindings of science and religious teach- considered Einstein’scriticism of the orthodox interpretation of quan-
ings and aspirations, be it Osiander’s preface to Copernicus’s treatise tum mechanics “high treason.” Being the official spokesman forthis
or community, he worried that too much dissent might harm its image
i the defense mounted by scientists against the accusations of theolog
i- and power. In an interview with Kuhn, Bohr complained aboutEin-
ans in the nineteenth century. Antirealism is especially valuable when
one wantsto protect science (and society) from undesirable extensi stein’s objections to the Copenhagen philosophy: “That, of course, is
ons really the difficulty . . . that the philosopher says: When the atomic
of scientific metaphysics into the political realm. Philipp Frank,i
n let- physicists do not agree, why should we trouble?” (17 December 1962,
ters to Bohr in 1936, tried to convince Bohr to join the camp of out-
spokenpositivists. The duty of every physicist, pleaded Frank, is to AHOP).
counter mystical interpretations of quantum physics, which were ex- Bohr’s antirealist philosophy was uniquely designed, not only to pre-
ploited by national socialist elements to support a reactionary vent discord inside the collective of quantum physicists, butalsoto pre-
political vent the alienation of this collective from therest of the scientific com-
philosophy anda barbaric regime. Onlya consistent positivist
stand munity. The appeal of Bohr’s teachings was considerably enhanced by
can insulate the words of physicists from such misuse (Heilbron
1987; his immense authority, personal charisma, and powerful institutional
Beller and Fine 1994, 20). .
Anantirealist approach canalso serve as a ‘potent cohesi position. Bohr’s status as a father figure goes a long way toward ex-
ve force in plaining the wide diffusion of the Copenhagenspirit (see chapter 12;
the scientific community. At the beginning of the century, Pierre
Du- Heilbron 1987). Yetit is the special role of Bohr’s antirealism as a source
hem argued thata positivist view of science, avoiding the minefie
lds of of cohesion that was largely responsible for the widespread appeal of
metaphysics, gives the greatest chances of achieving consen
sus in the the Copenhageninterpretation.
7. On the demonstration by Hipparchusof equiva
lent descriptionsof celestial motions
The new quantum physics was a theory of unprecedented abstrac-
using either epicycles or eccentrics, see Duhem (1991). tion, especially in its matrix and operator formulation. By renouncing
Poincaré’s leadership in the anti-
realist “descriptionistic’” movementis explored in Heilbr visualization in space-time, quantum theory challengedintuition (An-
on (1982). Convincing modern
arguments in favor of this view of science are made
conflict between scientific realism and the possibility
by van Fraassen (1980). On the schaulichkeit) and common sense even more severely than general rela-
of equivalent theories, see Ben- tivity had just a fewyearsearlier. The ascendancy of the profession of
Menahem (1990).
theoretical physics at the beginning of the twentieth century threatened
 178 Chapter Eight
The CopenhagenInterpretation and the Rhetoric of Antirealism 179
to upset the established hierarchy in the German physics community
(which placed experimentation andcollection of data at the top). Theo- of physics, especially those due to changes implied by the new
quan-
retical physicists seemed to alienate themselves even further from the tum theory. Physicists felt beleaguered by new “mathematicaltheori
es,
rest of the scientific community with the rise of the new quantum which are being continually formulated at an ever increasing tempo,
and in complexity and abstractness increasingly formidable.”
theory. The nonintuitiveness of the abstract matrix mechanics was Yet the
“average physicist... who flounders in bewilderment” need not
anything but appealing: Heisenberg faced the painful sight of Schré- feel
lost: “Since the new theories are formulated [so] as to be consist
dinger’s intuitive theory immediately enchanting mosttheorists. Schré- ent with
the cardinal principle that the properties of the thing have no meaning
dinger objected openly to the formidable complexity of matrix mechan-
which is not contained in some describableexperience, our intuiti
ics, warning that its lack of Anschaulichkeit would be detrimen on
tal to
scientific progress. I have argued that Heisenberg’s response—the
should be able to tell us what to expect in variousexperimental situa-
res- tions.” An analysis of thought experiments, using simple intuition,is
toration of space-time and Anschaulichkeit by an operational analysis
of open to every physicist and “would be found by many... . to give a
scientific notions in thought experiments in his uncertainty paper—
wasa clever defense of this highly abstract theory (chapters 2 and 4).
moreilluminating insight than a painful acquisition ofthe details of the
present mathematical picture” (Bridgman 1930, 21). Operational analy-
This operational analysis, despite Bohr’s initial objections in his Como
ses, of the sort advocated by Heisenberg and Bohr, were the anchor
lecture, became a paradigmatic examplein Bohr’s later legitimation and
and the chart that Bridgman offered to boost confidence in an age of -
elaboration of the quantum philosophy. uncertainty. Ss Po ; « &
Bohr’s numerous illustrations of experimental arrangements with Yet Bridgman’s work shows eloquently the dangers of operational-
bolts, springs, diaphragms, androds made experimentialists feel at ease
ism applied too diligently. Operationalism can undermine any well-
whenventuring into unfamiliar quantum territory. These simple consid-
established scientific theory, be it statistical thermodynamics or cos-
erations, supplemented with Bohr’s potent persuasive technique, which
mology, by attacking its foundations as going beyond the possibilityof
disguised demonstrations of consistency as compelling arguments of
experimental verification. Cosmologists, for example, resented Bridg-
inevitability, made the far-reaching and controversial conclusions ap-
man’s attacks on the integrity of their discipline (Beller 1988). It is this -
pear to be inescapable (see chapter 9). No mathematical training and
kindofoffense that Bohrcarefully and ingeniously avoided. The unique
only the most elementary physics was needed tofollow Bohr’s argu-
appeal of Bohr’s philosophy resided precisely in the wayit protected
ments. The development of the quantum theory, declared Bohr in one
the macrorealm and well-established classical theories from the ex-
of his numeroustalks, “has given rise to much anxiety, to much doubt
cesses of operationalism, while fully enjoying the fruits of antirealism
both among physicists as well as philosophers, and one {could] over-
in the microdomain. This situation was achieved byBohr’s division of
come these problems in a very simple way” (1958d, 726-27). Bohr’s
the physical world into two radically different parts—classical and
simple wayalleviated anxiety, transforming every readerinto a “virtual
quantum—and by Bohr’s doctrine of the indispensability of classical
witness,” to use Shapin and Schaffer’s (1985) term. It removed the threat
concepts. Before elaborating this claim, I will describe Bohr’s position
of esoteric and nonintuitive theory, making the quantum mystery

Likdae
as it contrasted with that of many mathematical physicists, including
widely accessible. English experimental physicists felt especially at
i 4

Born and Heisenberg.

home with Bohr’s drawings, so reminiscent of Duhem’s description of
an English factory. The English experimentalist Frederick Donnan was
Reality, Classical Concepts, and Symbols
enthusiastic: “As you point out, physical reality can only be defined by
te ERED

means of experiments and measurement... . Your logic—based on the

ALT nO

Oneofthe greatest philosophicaldifferences between Bohrand his more

complementary possibilities of different experimental arrangements
in mathematically minded colleagues concerns the questions of quantum
assigning definite and unambiguous meaningto the valuesof any
pair
RO Sais

reality and the necessity of classical concepts. Bohr transformedthefact

io

of canonically conjugate variables—seems unassailable. .. . It is not a

that one “can” describe a macroscopic measuring device by classical
Sormeecemampegeeraery ery opmuscesnopwes SPTLU

question of abstruse math but of an entirely newstandpoint” (Donnan

concepts (as opposed to a microobject, which cannotbe so described)
to Bohr, 26 July 1935, AHQP).
ie)

into the idea that one “has to”—into the doctrineof the indispensability
In the early 1930s, Bridgman described the anxiety and resentment
ofclassical concepts. Supplemented by the operationalstand,this doc-
causedin the physics communityby the rapid shifts in the foundations
trine resulted in Bohr’s peculiar denial of the possibility of quantum
Noe tee

180 Chapter Eight

The CopenhagenInterpretation andthe Rhetoric of Antirealism 181

concepts and quantum reality. According to Bohr, a symbolic descrip-

cal tools.” The word “gravitation” is an “artificial concept” that has
tion, removed from common sense and familiar classical visualization,
little in common with the simple idea—the feeling of force. Far from
can be nothingbuta tool for the correlation of experiments. Bohr’s doc-
being an extension of commonsense regularities, the laws of classical
trine of the indispensability of classical concepts rests on several key
mechanics “cannot be enunciated without ideas whichlie far outside
ideas that were not supported even by his closest collaborators—Hei-
the naturallimits of thought.”Justification of classical concepts is based
senberg and Born. Notorious among them are the alleged proximity
on nothing else but “their place in the system of objective naturalsci-
betweenclassical and commonsensenotions andthe direct accessibility
ence” (Born 1928, 21). Because the justification of classical concepts
ofclassicalreality tosense perception. ae ee, is due to their role in the abstract conceptual framework of classical
Bohr’s belief in this direct accessibility is rootedinthe Kantianheri-
theory, their ontological status is no different from that of the concepts
tage of space-time concepts as formsof intuition (Anschauung) and in
of quantum theory. There is no essential difference betweenclassical
‘his lifelong reliance on visualizable classical space-time models of the
and quantum concepts. .
atom as a basisfor correspondencearguments (on the former, see Faye
If reality is ascribed to the best entrenched theoretical concepts, they
1991 and Chevalley 1994; on the latter, Darrigol 1992a). The idea
that become“real” no matter how abstract. The exposure to general rela-
the classical world is revealed directly to our sensesis a recurrent theme
tivity weaned physicists from their familiar classical intuitions and de-
in Bohr's writings. This direct and necessary connection between clas-
veloped suchrealistic symbolic sensitivities. Einstein and his follow-
sical concepts and sense impressions ensures the absolute indispen-
ers ascribed genuinereality to four-dimensional space-time and tensor
sability ef classical concepts: “There can, of course; be no question...
fields, not to familiar classical space and time notions. Despite the de-
of abandoningthe conceptual frame of our conscious recording of sense
manding symbolization, these concepts becamereal for physicists by
impressions” (Bohr 1938a, 379). Bohr’s intuitions about. whatis “real”
being included into a new comprehensive and consistent theory. And
and whatis not are connected with the possibility of visualizing con-
conversely the regular notions of space and time becamelessreal: “The
ceptual notions, not with the possibility of observing them. Atoms are
familiar geometry of Euclid and the corresponding time are now re-
therefore “real’”—thecalculation of their masses and charges is based
duced to mere approximationto reality” (Born 1928, 25). Similarly, Hei-
on the visualizable lawsof classical electrodynamics.
senberg, redefining Anschaulichkeit in his uncertainty paper, argued that
~ Yet when wedealwith the unvisualizable symbolic description of the
the simplest interpretation of space coordinates in quantum theory is
same atom, we should leaveall questions of its “nature” behind. Be- obtained by identifying them with the diagonal terms of the appropri-
causeclassical concepts are inapplicable, or limited, in the quantu
m do- ate matrices.
main, and only classical concepts are anschaulich and therefor
e “real,” Heisenberg’s published works, and the reminiscences of his student
there can be no question of “quantum reality.” There is no compelling Weizsacker,testify that Heisenberg did not embrace Bohr’s radicalsplit-
argumentforthereality of classical description as opposed to quantum
ting of the world into two unbridgeable realms: “Heisenberg neverse-
description, exceptthis alleged Kantian kinship betweenthevisualiza- riously considered the idea that quantum theory might be limited to
bility of classical physics and our sense perceptions.
microscopic objects; . . . he considered quantum theory as universally
According to Bohr, commonsense conceptionsand ordinarylanguage
valid and classical physics as macroscopic approximation” (Weizsacker
are the foundation ofall communication, and therefore ultimately
of 1987, 283). This willingness to reeducate their intuitions and to rede-
all science. According to Born and Heisenberg, common senseis pro-
fine reality characterized mathematical physicists working in quantum
foundly mistaken. In Born’s opinion, the guiding factor in the develop
- theory. Far from believing that no new conceptual schemesofreality
mentof science is “a belief in a real external world and an ability
to are possible, Born held that the development of such schemesis the
distrust sensation’ (1928, 20). Born held that commonsense and
scien- very essenceofscientific advance. In developing new invariants, phys-
tific concepts are greatly removed from each-other. The concepts
of icists learn to handle them intuitively. As a result of such a process,
classical mechanics are not extensions of common sense. Thefact
that “new conceptions sink down into the unconscious mind, they find
classical physics uses such everyday terms as “mass” and
“force” is adequate names,andare absorbed into the general knowledge of man-
misleading: “Their sound is the same as words of ordinary speech,
kind” (Born 1936, 52).
yet their meaning can be found only from specially formula
ted defini- Bohr, being a “natural philosopher, and not a mathematical physi-
tions,” and they can be apprehended “only with the aid of mathema
ti- cist” (Heisenberg’s characterization), did not possess the experience or
 182 Chapter Eight
The Copenhagen Interpretation andthe Rhetoric of Antirealism 183
dexterity needed to handle the abstract mathematical formalism of
gen- describe the general laws actually now followsthescientific language
eral relativity. In fact, his familiarity with general relativi
ty was very of the mathematicians” (Heisenberg 1958, 172).
limited (Pais 1991). His objections to the possibility of
new conceptual Although few working physicists abided by Bohr’s strictures
schemesare peculiarly reminiscentof the objections on
of the neo-Kantian quantum concepts and quantum reality, their disagreement with Bohr
opponentsto general relativity. Bohr did not internalize
the new quan- rarely reached the printed page. Yet the disagreement was often un-
tum symbolism as others did who made the necessa
ry investment in equivocal and forceful. Bohr’s correspondence with his lifelong col-
everyday mathematical labor.* For him, only the classica
l realm re- league Born and with the Russian physicist Vladimir Fock provides
srl

mained intuitively accessible and therefore real.

characteristic examples. Fock attacked thecentral points of Bohr’s phi- ©
In his later years, Bohr replaced his initial, Kantian emphasis on
the losophy inhis letter to Bohr of 23 February 1957 (AHQP), andin a
impossibility of new conceptual forms with a different argume
nt for paper published the next year (Fock 1958). Both Fock and Born were
the indispensability of classical concepts. This argument was
based amazed by Bohr's rejection of quantum concepts andfrustrated with .
on Bohr’s post-EPR use of operationalism and onhis
segs.

doctrine of the his resignation from the attempt to comprehend the underlying quan-"
unambiguous nature of communication in terms of
common language. tum reality. Fock argued: “It is not only the limitations characteristic
Bohrheld that onlyclassical language, as a natural extension of com- of a description of phenomena. . . that have philosophical significance,
mon language, is capable of communicating physical experience. Ac-
but also the constructive part of quantum mechanics and thenew fun-
Hest Re REINERTINTSTL

cording to Bohr, the unambiguous communication of measur

ement damental conceptsconnected with it. ... The brilliant demonsiration
results is the.only possible way to ensure the objectivity of the scien-
given by Bohr ofthe limitation of classical concepts is not accompanied
tific enterprise (a penetrating analysis of this position can be
found in by even theslightest indication of new concepts by which to replace
Hooker 1972, 1991). Complementinghis instrumentalist insisten
ce that them” (1958; quoted from Fock’s translation, AHOQP). Similarly,Born
quantum formalism is merely a tool for coordinating measur
ementre- claimed that science aims beyond mere coordination of measurement
sults, Bohr declared the classical realm the only one whererealistic
talk readings, that there is something objective “behind the phenomena”
can be meaningful. Any creative efforts to constructthe quantum world
(Born to Bohr, 28 January 1953, AHQP). Unless one takes a realistic
become pointless andare boundtofail. Yet the doctrine of the
indis- stand, one “resignsoneself to answering any question about whyoneis
pensability of classical concepts for the unambiguous
description of investigating the [world] at all” (Born to Bohr, 10 March 1953, AHQP).
measurement does not by itself prohibit the construction
of a Fock disagreed vehemently with Bohr’s opinion that the quantum
quantum reality. An emphasis on experimentis not at odds with
realist formalism was merely a tool for the coordination of measurementread-
strivings, as is eminently displayed in writings by
Born, Heisenberg, ings. According to Fock, such a point of view is erroneous: “True
and Pauli. As Born noted: “In physics, all ‘experience’ consists
of the enough, quantum mechanics(like any other physical theory) allowsus,
activity of constructing apparatus and of reading instrum
ents. Yet the amongother things, to coordinate the readings of the instruments used
¥ it ARSE

results thereby obtained suffice to re-create the cosmos by thought”

for the measurement, but that is not its fundamental importance. The
(1928, 36). . purpose of a physical theory is always to describe the properties of
Nor did Bohr’s mathematical colleagues share his opinion that
un- physical objects in their relation to the outer world” (Fock to Bohr,
ambiguous communication demands the use of commo
n language. 23 February 1957, AHQP).
Quite the opposite stand is expressed by Heisenberg:
The concepts of Fock rejected Bohr’s idiosyncratic claim that the “mathematical sym-
ordinary language are “inaccurate and only vaguely defined
... . the bols of quantum mechanics, in contradistinction to the mathematical
concepts of the general laws must in natural science be defined with
complete precision, and this can be achieved only
symbols ofclassical physics, have no physical meaningin themselves.”
by means of mathe- He did not see any difference in the role played by mathematicsin clas-
matical abstraction.” Far from immutable, scientific language evolves
sical and in quantum mechanics. The notions of quantum state and
as scientific knowledge grows: “New termsare introdu
ced and the old quantum probability are in fact fundamental quantum concepts. “They
ones are applied in a wider field or differently from
ordinary lan- are not symbolic, but quite physical. There is no reason to avoid them
guage. ...in the theory of general relativity the language by
which we in the description of nature” (Fock to Bohr, 23 February 1957, AHQP)?
8. On Bohr’s attitude toward mathematics, See chapter 12.
9. Similarcriticism has been provided by Krips (1987).
sciste tance tae
184 Chapter Eight
The CopenhagenInterpretation and the Rhetoric of Antirealism 185
i The Appeal of Antirealism: Bohr’s Version ficulties remained, as I argued in the previous chapter. Heisenberg
Hi
gi grasped that the operational analysis of quantum phenomena through
Antirealism in Bohr’s philosophy was associated with the unfamiliar the useof classical terms exclusively cannot be contradiction free (Beller
and (according to Schrédinger) “frightfully abstract” quantum mathe-
1993). Other physicists expressed similar reservations in unpublished
matical formalism, not with the classical realm knownto every physi-
correspondence with Bohr (for example, Fritz Bopp’s letters to Bohr,
cist. The proximity between commonsense notionsand classical con-
AHOP). Yet in print not many dared,or cared, to enter the territory that
cepts, the direct accessibility ofclassical reality to sense perception, the only Bohr, so the legend goes (Rosenfeld 1961; Blaedel 1988), had mas-
ANRCSRT eter

need to use commonandclassical language for unambiguous commu-

tered fully.
nication, the needto describe measuring apparatus using classical con-
cepts—Bohrpresentedthemall as self-evident truths rather than as the
Antirealism and Opposition
controversial doctrines they, in fact, were. And their appeal to experi-
mental physicists was very strong. Born understood it clearly: “Every
Ihave arguedthatscientific philosophizing in general, and that dealing
_ experimental physicist treats his instrumentation as if he were a naive
with the realism-positivism issuein particular, is most profitably un-
realist. He takesits reality as given a priori, and doesn’t rack his brains derstood by juxtaposing the universal “must not” with the local “need
about it. Niels Bohr has madethis attitude the basis of his whole phi- not.” Byits very nature, “need not” brings us to a specific historical
losophy of physics” (1962, 21).
epeeneenare ee li

situation, to local argument designed to dispose of deeply entrenched

Born did not share Bohr’s views. Yet Born understood the immense but no longer useful habits of thought, to an appeal addressed to a
potential of Bohr’s stand as a unifying force. So Born would sometimes
specific audience. While philosophically questionable,it is rhetorically
Sanne

endorse Bohr’s radical split of the physical world in order to avoid a wise to endowa local “need not” with universal validity. Attempts
split in the world of physicists: “There has: already developed a gap to mesh different coriceptual strands into a systematic position lead to
GPa RSPAS are

between pure and applied science and between the group of men de- such philosophical hybrids as Murdoch’s characterization of Bohr as an
voted to the oneor the other activity, a separation which may lead to “Instrumentalistic realist” (1987, 222). It is not surprising therefore that
a dangerous estrangement. Physics needs a unifying philosophy, ex-
merge ppneverescaeuivnns ae ee

whatis called the Copenhageninterpretation is so riddled withvacil-

pressible in ordinary language, to bridge this gulf between ‘reality’ as lations, about-faces, and inconsistencies. For quantum philosophy—
thoughtof in practice and in the theory” (Born 1953c, 151). “Plain,” or which was, moreoften than not, elaborated asa series of local responses
“ordinary,” languageis the only kind the expert can useto tell the wider to challenges from the opposition, rather than as an attemptto construe
audience aboutnewabstract technical advances: “Thephysicist may be a systematic framework—by the very nature of its development cannot
satisfied when hehas the mathematical scheme and knows how to use
Y

be foundto be free of contradictions.

it for the interpretation of the experiment. But he has to speak about Manyof the antirealist statements of the Copenhagen camp camein
his results also to non-physicists whowill not be satisfied unless some response to Schrédinger’s and Einstein’s critiques of quantum theory.
explanation is given in plain language, understandable to anybody”
oe Lg ee

Asthe nature and forcefulness of the opposition’s threat changed,dif-

(Heisenberg 1958, 168). Heisenberg’s writingsare full of contradictions
ferent, often contradictory claims on the same issues appeared in writ-
precisely because he simultaneously addressed two “complementary”
ings by Bohr, Heisenberg, and Born. Opinions aboutthe “reality”of the
audiences—thelaity and the mathematically initiated. : y-function constitute a prominentcase in point. At the end of the 1920s,
Bohr’s doctrineof the necessity of classical concepts worked wonders
when Schrédinger, hopingto eliminate quantum jumps, attempted to
to alleviate the anxietymanyphysicists felt about symbolic quantum
substitute a wave modelfor the Géttingen-Copenhagenparticle model,
theory. His operational analyses of thought experiments were both a
his efforts were discredited by reference to the multidimensionality of
therapeutic compensation for the loss ofAnschaulichkeit and a crutch
the wave function. Because the ¢-function was multidimensional, the
(London's term; Landé 1967), thougha brittle one, on which scientists
matrix physicists argued, it was not something “real,” but rather an
could lean while familiarizing themselves with the quantum mystery.
abstract notion, having great mathematical utility but no physical sig-
Howbrittle the crutch was soon became apparent because of inconsis-
nificance. This argumentwasclearly targeted at an audience whoiden-
tencies connected with Bohr’s and Heisenberg’s initial ideas about the
tified anschaulich with “real,” in the neo-Kantian spirit.Such an identi-
disturbance of atomic objects by observation—the notion Bohr even-
se Sr

fication wasat odds with the more sophisticated notionsof Heisenberg

tually had to renounce. But even after his renunciation, insoluble dif-
and Born, who viewedreality as flowing from features of the mathe-
Sry i

186 Chapter Eight

The Copenhagen Interpretation and the Rhetoric of Antirealism 187
matical mode]. Heisenberg and Jordan’s pronouncements on the unre-
mathematical manipulation. He contrasted his scientific approa
ality of the y-function exasperated Schrédinger. In vain, Schrédinger ch with
that of a layman, for whom “the concept of reality is too much
appealed to Born, who backed his younger colleaguesfully onthis issue con-
nected with emotions to allow a generally acceptable definit
(chapter 2). The possible criterion of reality by which Born considered ion” (1964,
103). Yet perhaps the difference between a layman and a quant
particles real, yet the wave function abstract, was a mystery to Schré- um
physicist in this respect is not as great as Born claimed. Onecou
dinger. He explained to Wien: “I believe that Born thereby overlooks ld say
that for scientists, as for laymen, “the real things are those
that . . . it would depend on the taste of the observer which he now things which". -
are importantfor them”(1964, 103). z
wishes to regard as real, the particle or the guiding field. Thereis cer- ”
tainly no criterion forreality if one does not want to say: thereal is only
The Appearance of Consensus and Conclusion |
the complex of sense impressions, all the rest are onlypictures” (quoted
in Moore 1989, 225).
In reading philosophical accounts written byscientists,it is important —
Whentheoriginal challenge posed by Schrédinger’s interpretive ef-
to explore fully the contradictions therein, rather than selecting and
forts faded, Born changed his presentation of the significance of the w-
constructing an artificial, consistent system of belief. Forit is precisely
function.Ina lecture given in 1950 Born asked: “Whatabout the waves? those claims that refuse to be neatly systematized that are the most il-
Are they real and in what sense? . . . Though the wave-functions are luminating. Idiosyncratic contentions, puzzling stands, seeming rever--
representing, bytheir square, probabilities, they have a characterof re- sals—these are the most useful guides to the fascinating local context,
ality. That probability has some kind ofreality cannot be denied. How so different from the rational reconstruction that ultimately serves the
could, otherwise, a prediction based on probability calculus have any interests of the orthodox.
application to the real world?” Born concludedthat “the scientist must a
Full acknowledgmentof its contradictions casts great doubt on the
be a realist” and that the abstract notions of a mathematical formalism
existence of a commonbasis for the Copenhageninterpretation, and
can—and must—beassigneda realistic significance: “He[thescientist] perhaps on the notion of a conceptual framework in general. The cen-
usesideasof very abstract kind, group theory in spacesof many or even
tral pillars of the Copenhagen philosophy are indeterminism, wave-
infinitely many dimensions and things like that, but finally he has his - particle duality, and the indispensability of classical concepts. Yet none
observational invariants representing real things with which he learns of them commands unwavering commitment even from Bohr’sclosest
to operate like any craftsman with his wood or metal” (Born 1950, 106).
collaborators. Nor did these concepts have unchanging and definite
Three years later Born made another about-face in his evaluation of
meanings for Bohr himself.° Born and Pauli, as their correspondence
the reality of the w-function. Schrédinger (1952a) had renewedhis at-
discloses, did not think that the bankruptcy of determinism could be
tack on the Copenhagen interpretation and revived his old proposal to compellingly deduced; neither did Born and Heisenberg fully share
eliminate quantum discontinuities and irrationalities by using an in-
Bohr’s understanding of wave-particle complementarity; nor did
tuitive wave model. Born responded immediately with a counterattack Bohr’s mathematically skilled colleagues adhere to his dictum of the
that wasa repetition of the argument against Schrédinger madethirty- indispensability of classical concepts, as I have argued.It is difficult to
five years before: “A multi-dimensional wave function is nothing but a

&
find any common denominator, any idea, that enjoyed apparent “com-
tah
nameforthe abstract quantity of the formalism, which by someof the
dda mitment” that did not also see that “commitment” brokenat one time
modern theorists also goes under the more learnedtitle of ‘state vector
or another. This is not to say there was no consensus that the adventof
in the Hilbert space’” (1953b, 143). Born repeated his strong commit-
mentto the reality of particles as opposed to waves. Thus the assign-
mentof reality or lack thereof can be:understood properly only in the 10. The opposition also lacked a coherent response. Fine (1993) argues that Einstein
local context of a controversy; realism or antirealism regarding the subscribed to three distinct kinds of interpretations of the quantum theory—subjective,
instrumental, and hidden variables— giving himself considerable interpretive leeway.
same theoretical construction is often a rhetorical weapon, whosesig- 11. “It would besilly and arrogant to deny any possibility of a return to determinism”
nificance can only be grasped within the definite sociohistorical circum- (Born 1964, 108). Heisenberg, as he revealed in an interview with Kuhn, neverliked
the
stances in whichit is used. “dual” complementarity approach. “We have one mathematical schemethatallows many
Attempting to present an “objective” definition of whatis real, Born ‘transformations’: The fact that we can use two kinds of words {“wave” and “particle”
]
(1928, 1964) equated real things with invariants of observation and is just an indication ofthe inadequacy of words” (interviewwith Heisenberg, 25 February
1963, AHQP). See chapter 11 for a full discussion.
 188 Chapter Eight The Copenhagen Interpretation and the Rhetoric of Antirealism 189
quantum theory necessitated a radical departure from classical physics. Wave-particle complementarity is another prominent case in point
It is exact and consistent agreement about what would constitute such (for further discussion, see chapter 11). Bohr argued that one “has to”
a departure that was lacking. The bankruptcy of the classical idea of use both wavesand particles in quantum mechanical description, pre-
motion wasas fully grasped bythe opposition—Schrédinger and Ein- senting wave-particle complementarity as an epistemologicalnecessity.
stein—as by the orthodox quantum physicists. It is true that Einstein Born and Heisenberg held that one “can” describe quantum mechanical
had serious reservations about quantum theory, andit is also true that phenomenausing both notions, andthey often did so. Neither believed
Bohr andhisfollowers were ready to accept the finality of quantum that one “has to” employ both waveandparticles for a comprehensive
theory, no matter whatfar-reaching philosophical revisionsit might im- description of every quantum phenomenon, nor did Born and Heisen-
ply.. Butfinality is an ideological, not a conceptual, position, andthisis berg hold that these concepts are necessarily mutually exclusive. Nei-
_ pethaps the reason scientific controversy often looks morelike a polit- ther waseager, of course, to publicize his disagreementwith Bohr.
ical campaign, with one side discrediting and caricaturing the other, In the early 1930s Born was lukewarm to Bohrian elaborations of
than an open-minded dialogueabout fundamentals. complementarity. In 1929 he claimed that the resolution of the wave-
Nowwehave to face the question: How does the impression (illu- particle dilemmaresulted first and foremost from his ownstatistical
sion) of the integrity, of the solidity, of a conceptual frameworkarise? interpretation: “If we disregard all philosophical aspects, the contradic-
By whatrhetorical means are we led to believe that there is more sta- tion between the corpuscular and wave properties of radiation would
bility in the views of acertain scientist, and more consensus among be insoluble withoutthis statistical viewpoint” (Born 1928, 34). Born’s
different participants, than is actually the case? The question about the philosophical discussions in the 1930s were concerned with indeter-
viewsofan individual scientist is perhaps the easier of the two:In or- minism and hardly mentioned complementarity. In the 1950s, the pub-
der to persuadeothers or to appéar credible, scientists often present lic image of physics and the status of physicists underwenta drastic
changes in their views as natural elaborations of previous positions, change, due to the use of the atomic bomb in-Japan—a disaster that
rather than as major revisions or reversals. Substantial changes, as we affected Born deeply (Born 1951). Born subsequently embraced the
have seen in previous chapters, unacknowledged by the authors, are complementarity philosophy because of what he perceived as its con-
characteristic of the philosophical thought of Heisenberg, Bohr, Pauli, ciliatory message: “The world which is so ready to learn the meansof
and Born. Andthere is nothing distinctive here about quantum physi- mass-destruction from physics would do better to accept the message
cists—to give just one example, Newtonian scholars have found that of reconciliation contained in the philosophy of complementarity”
Newton,in the controversyoverthe nature of white light, similarly pre- (1950, 108). Considering complementarity a “healthy doctrine,” suit-
sented what wasa forced departure from a previousstand as merely an able to “remove manyviolent disputesin all waysoflife,” Born insisted
elaboration of it (Shapiro 1989). that the “real enrichment of our thinking is the idea of complementar-
The more difficult question is how the appearance of consensus ity” (1950, 107).
amongdifferent participants is achieved. Here againthe distinction be- Mypresentation ofantirealism in theinterpretation of quantum phys-
tween “need not” and “must not”is helpful. Closely related to the issue ics distances itself from such notions as “belief,” “commitment,” and

2cvsbe dS
of “need not” versus “mustnot”is the question of what one “can” and “metaphysical influence” and denies the very possibility of presenting
_ whatone “has to” do. Failure to discriminate between the twoleads to the Copenhageninterpretation as a coherent philosophical framework.
the impression ofa firmer consensus between quantum physicists than Whatever cohesionscientific paradigms or conceptual frameworks ex-
actually existed. What Heisenberg and Born often promoted as a pos- hibit has to do more with the integrity of the tools (mathematical,
sible epistemological stand, Bohr advanced as an inevitable one. In the experimental) than with “shared” metaphysical presuppositions.2 We
fact that one “can” describe a macroscopic measuring device byclassi-
should beware of conflating “exposure to” (a certain viewpoint) with
cal concepts, Bohr transformed “can” into “has to,” elevating
the fact “influence,” of hastily translating “familiarity” into “predisposition”
into the doctrine of the indispensability of classical concepts. In con-
trast, Heisenberg, more often than not, did not say that one “has to” 12. Oneinteresting attemptto find somestability, objectivity, and cohesion, despite the
use classical concepts—rather his claim is that we are “satisfied” to use impressive impactof social studies of science that deny those characteristics to science, is
classical concepts, or that we “feel entitled” to use them (Heisenberg that of Ian Hacking (1992b). Hacking sees stability and objectivity in developed “stylesof
reasoning” that, proving their value and eventually shaking off their social origins, be-
1938). come universal.
 ALCS SE Te aE

190 Chapter Eight

and “predisposition” into “commitment.” The pointis not merely

that
it is exceedingly difficult to formulate a consistent framework
even
when oneintends to do so.It seems to be impossible to reconst
ruct a
coherent philosophical framework from a multitude of utterances and
deliberations that were aimed at meeting challenges in shifting scien-
pennant, epee

tific and sociopolitical circumstances.

13. “We characteristically spill over the limits of our intelligence . ..

and get confused,
and.. . attempts to synthesize our views may in consequencereveal conceptu
al disorder
at least as muchas coherent doctrines” (Skinner 1969, 49).
0-70

i
Et
:

Above, According to the received history, Werner Below, The dialogue between Bohr (left) and Al-
Heisenberg(left) and Niels Bohr (right) came to bert Einstein (right) lasted from theearly 1920s
complete agreementafter their heated debates in until the end of Bohr’s life. While it is often as-,
1926 overthe contentof Heisenberg’s uncer- serted that their debate over EPR was a mile-
taintypaper. Yet an analysis of their dialogue re- stone in Bohr’s “victory” over Einstein concern-
veals that despite a commonpublic stand, the ing interpretive issues of quantum theory, a
disagreements between them were never re- critical analysis undermines such a reading (see
“solved (see chapters 6, 8, and 11). chapter 7).
coei
SeRamnen
IeMoen eet EY. AVES
 Above, Heisenberg (right) and Above, The experimental work
Wolfgang Pauli (center) corre- of James Franck (center) pro-
sponded frequently during the vided crucial supportfor the
crucial years of the creation of concepts ofstationary states
the new quantum mechanics and quantum jumps proposed
andits interpretation. Heisen- by Bohr (left). The issue of
the
berg wrote his uncertainty pa-
“reality” of quantumjumps
per in an intense dialogue with wasat the center of interpretive
Pauli (see chapters 2 through 5)
‘attempts between 1925 and
Otto Stern appears at the far 1927 (see chapters 2, 4, and 6).
left.
Cate a rereene

Hans Hansen appearsat the

right.

Left, Heisenberg (right) and Er-

Right, Paul Dirac(left) and Hei-
win Schrédinger (left) shared
senberg(right) in the 1930s. The
the Nobel prize in physics for
dialogue between Heisenberg
their matrix and wave versions
and Dirac wascrucial for the
of quantum mechanics, Tespec-
development of quantum me-
tively. Heisenberg’s inner dia-
chanics andits physical inter-
logue with Schrédinger was in-
pretation in general, and for the
dispensable for the formulation
emergence of Heisenberg’s un-
of the uncertainty Principle
certainty principle in particular
(chapters 2 and 4). The king of
(see chapter4).
Sweden appearsin the center.
 Left, Experimental work on
atomic collisions by Franck
(right), Max Bom’s(center) col.

Pow
Below, Pauli (right) was one ofthe
main architects of the Copenha-
gen interpretation. Bohr's (left)
dialogue with Pauli was instru-
mentalin the initial formulation :
andlater dissemination of com- 3
plementarity (see chapters 6
and 12).
ere
Pet
ee
Ree Ips
coe):BabarNNW:
i

As Bohr’s dialogue with Einstein(left) lasted a lifetime, so did Pauli’s (right). Pauli‘’s predilection for
ne

the philosophy of complementarity was reinforced in his later years by his spiritual and metaphysi-
cal longings (see chapter 12).
j
2
;

A dialogical analysis unveils the importance of “lesser”scientists, whose names rarely appearin
the received history of quantum physics. Frits Zernike’s work on thelimits of precision in measure- The work of the American physicist William Duane, whousing corpuscular light quanta explained
ments spurred Heisenberg’s deliberations on the theoretical uncertainty in quantum mechanics (see the diffraction of light by a grating, stimulated theinitialtreatment of photons and electrons purely
chapter4). as particles in Heisenberg’s uncertainty paper, as well as Alfred Landé's later particle interpretation
(see chapters 4 and 11).
 CHAPTER 9
XZ
The Copenhagen Dogma: The Rhetoricof
Finality and Inevitability

You cannot revolt:this is two times two equals four! Nature does not ask you,
she cares nothing about your wishes and whether you like or dislike her laws.
You must accept the way Nature is.... A wall, this means, is a wall... and
soon...andsoon..:
Fyodor Dostoyevsky, Notesfrom the Underground. mytranslation

Makeit look inevitable.

Louis Pasteur’s advice on the best strategy of persuasion, quoted in
Holton 1991, 174

Tf it is so, it must be necessary,” was a remark which he [Bohr] would often

make in a discussion, and he would notbe satisfied until the necessity was
completely elucidated.
Jorgen Kalckar 1967, 233

Introduction
The founders and followers of the Copenhagen interpretation advo-
cated their philosophy of physics not as a possible interpretation but
as the only feasible one. Attemptsat basically different approaches,al-
beit by such prominentscientists as Einstein, Schrédinger, Landé, and
Bohm, were dismissed and ridiculed (see chapter 13). Niels Bohr ad-
vanced the Copenhageninterpretation relentlessly. He published ex-
tensively, lectured widely, and spread the Copenhagenspirit in direct
contact with a generation of physicists who visited the Institute of
Theoretical Physics in Copenhagen (Heilbron 1987). Bohr wasa skill-
ful rhetorician of the quantum revolution, no less ingenious than Gali-
Bohr (left, facing camera) in a discussion with Pauli (left, back leo in weaving “natural interpretations’—arguments designed to in-
to camera), Lothar Wolfgang Nordheim
(next fo Pauli), Leon Rosenfeld (next to Bohr), and an unidentif
ied scientist. During Bohr’s professional troduce controversial knowledge in the guise of intuitively appealing
life, the Copenhageninstitute attracted physicists from
around the world. Heisenberg and Dirac propositions.?
spent the winter of 1927 in Copenhagen—both Heisenber
g’s uncertainty paper and Dirac’s paper on
transformation theory were published from Copenhagen (see
chapter4). 1. The term “natural interpretation” is due to Paul Feyerabend (1975), who analyzed
Galileo’s rhetoric using this notion.
 192 Chapter Nine
The Copenhagen Dogma 193
There is, no doubt, a striking contrast between the founde
r of com- infects the reader with the sheerintensity of Bohr’s conviction,is skill-
plementarity and chief spokesmanof the quantum revolution
ora Na eet

and that fully interwoven with the intimidating rhetoric of Bohr’s authority.
of the Galilean. Galileo—passionate, impatient, and breathtaking
ly elo- Surely, an intelligent reader cannotfail to realize, intimates Bohr, some-
quent; Bohr— desperately struggling for words and talking
in an “in- thing that is. “obvious,” that is “evident,” that was “clear from the out-
audible manner” (Heilbron 1987, 201), or more reverently,
a “divinely set,”"and thatis, after all, “a simple logical demand” (1937a, 1948).
bad speaker” (Pais 1991, 11). Galileo was superb at antagonizing
his In addition to these forceful strategies of persuasion are others, more
audience, Bohr skillful in creating a favorable mood.
Bohrrarelyfailed subtle. One of the most potent consists of Copenhagen’s arguments of
to praise the achievements ofhis listeners or to mention the
PNT:

contribu- “fnevitability,”, which are, as I have mentioned, merely disguised ar-

tions of the host institution. The kind of audience determined the
way guments of consistency. The opposition’s resistance to this technique
Bohrpresented his epistemological efforts—as “philosophical
”in a re- of argumentation, and its subsequentrejection of the Copenhagen phi-
ceptive Danish setting, or as those of a pure physicist, anxiou
s not to be losophy, is therefore reasonable, and not “conservative” as is often
identified with philosophers, in a more pragmatic Americ
an environ- ‘claimed by adherents to the Copenhagen orthodoxy. The Copenhagen
ment: “I am very muchafraid I shall appear to you as a philos
opher. ... physicists used their arguments of inevitability to show the impos-
I am just a physicist who has been forced to enter on such proble
ms” sibility of observer-independent, deterministic description in the quan-
(Bohr 1937d, 264). The word “forced” contains a subtle vet
powerful tum.domain. The rhetoric of inevitability was especially disturbing, as
rhetorical message. Not merelydoesit legitimate Bohr’s
continuing pre- J. S. Bell mentioned, after the 1950s, when David Bohm provided fre-
occupation with philosophical matters, it endows that preocc
SeSaRRCECerenga a

upation cisely the sort of description that the Copenhagen physicists had sup-
with inescapability and Bohr’s conclusions with an aura of inevita
bility. posedly: ruled out “in principle” (see Cushing 1994b for an excellent
Einstein characterized the young Bohr as a thinker who “utters
his study of Bohm versus Copenhagen;see also the collection of papers in
opinions like one perpetually groping and neverlike one
who believes Cushing, Fine, and Goldstein 1996).
he is in possession of definite truth.” Pais (1991) chose these words
as Referring to the orthodox denial of the possibility of a quantum on-
an epigraph to his book on Bohr. Whilethis portrayal might be
fitting tology, Bell invoked a vivid image:“It is,” he wrote, “as if our friends
to describe the young Bohr’s attempts to cope with the quant
um mys- could not find wordstotell us aboutthe very strange places where they
tery, it is hardlytrue of the atmosphereof Bohr's later diffusi
on of the went on holiday. We could see for ourselves whether they came back ~
Copenhagenspirit. The later Bohr spoke as a prophet of truth,
with browneror fatter” (1987, 171). Why were quantum physicists mute for
the intimidatingprivilegeof ultimate authority, in a categor
ical manner so long on the issues of quantum reality? Whydid they voluntarily give
that excludes, “in principle,” any questioning. Bohr‘s writings are
thick up an exciting, though admittedly inconclusive, quest to unveil the
with such expressionsas “we see that it cannot be otherwise,”
“this is mystery of the quantum world? Whydid they hold fast to the notions
something there is no way around,” “something we believe is
going to of observer dependence and inevitable acausality, despite the existence
stay forever,” “the situation is an unavoidable one”
(1935b, 210, 215). of Bohm’s alternative version? Bohm attributed this puzzling situation
Bohrpresents complementarity as “the most direct expres
sion of a to the inability of the orthodox physicists to conceive of any concepts
fact . . . as the only rational interpretation of quantum
mechanics” other than classical ones (waves and particles), and to their commit-
(quoted in Landé 1965, 126), founded on an “obvious”
analysis of the ment to the philosophy of positivism: “There appears to have existed,
“epistemological situation.” In viewof such a compelling
“logical” especially among physicists such as Heisenberg, and others, whofirst
analysis, complementarity is an inevitable, irrefutable, final
principle of discovered the quantum theory, a rather widespread impression that
physics, from which “we cannot deviate, whatever progre
ss the future the human brain is, broadly speaking, able to conceive of only two
may bring” (Bohr to Born, 2 March 1953, AHOQP). Bohr’s
later writings kinds of things, namely fields and particles” (1957, 96). Another reason
are especially dense, sentenceafter sentence studded
with such expres- for their refusal to consider a “subquantum mechanical level,’” accord-
sions as “outof the queStion,” “inno way,” “of course,
” “surely,” “leaves ing to Bohm, wasthe adoptionofthe positivist thesis of the elimination
us no choice” (1937a, 1938a). Bohr’s arguments are often prefac
ed with of unobservables from physical theorizing, “a general philosophical
such phrases as “we must recognize,” “we must
realize,” “the main point of view containing various branches such as ‘positivism,’ ‘op-
point[to] realize,” “it is importantto realize,” “itis
imperative to real- erationalism,’ ‘empiricism,’ and others, which began to attain a wide-
ize” (1930, 1937a, 1938b, 1948). This strategy
of persuasion, which spread popularity among physicists during the twentieth century”
 194 Chapter Nine
The Copenhagen Dogma 195
(1957, 97). I will reiterate Bohm’s insight about the intima
te interrelation between twospecified values (chapter4). In this way discontinuity and
among the finality of acausality, the indispensabi
lity of classical con- acausality were entrenchedin the very axiomatic basis of quantum me-
cepts, and positivism but provide a different account
of how they are chanics to such a degreethat they appeared to be “hard facts” of nature.
connected, along more sociologicallines.
-From a certain perspective, the stand that quantum This reading dispels two widespread misconceptions concerning the
theoryis funda- history of quantum mechanics. Thefirst misconception is that an ab-
mentally acausal can seem natural and legitimate. If
our best scientific stract quantum formalism developed before any interpretive concerns
theory is ipso facto acausal, what reason, beyon
d an unwarranted a
Priori belief in classical causality, would urge physicists were raised, simply as aresult of disinterested mathematical activity.
to search for According to the second misconception, the Gottingen-Copenhagen
a causal version? After all, many physicists, including
Bohm himself, interpretation ‘camefirst, and only later did Schrodinger begin his
found determinism neither indispensable nor appeal
ing. Heisenberg’s competing attemptsat interpretation. Both claims, we have seen, must
argument—thatif physicists are convinced of the correc
tness of quan- be qualified substantially. From the beginning, the orthodox inter-
tum theory, which is built on probabilistic foundations,
then they pretation and the formalismitself were developed to defend the past
‘should accept acausality (1958, 89)—is a characteristic
example of a le- achievements that Schrodinger threatened to undermine and, through
gitimate kindof reasoning that ArthurFine called “entheorizing” (1986,
its philosophical legitimation, to protect the Gatlingen: opensagen
87). The metaphysical question of whether a certain property
is a prop- research program against any challenge.or alternative. The am ni
erty of our world is translated into the question of whether a
theory Heisenberg, wholaid the foundationof matrix mechanics, was one .
that entails this propertyis fruitful, or correct. Yet such sober
entheoriz- the central players in this enterprise. Strong sentiments against ae
ing is rare. Copenhagenwritings, Heisenberg’s included, are laden
with ity quickly arose precisely because causality was an intrinsicpart of the
the heavy rhetoric offinality, of the inevitability of acausality,
of the the opposition. -
impossibility of a causal alternative. An analysis of the Copen
hagenar- atoneeetfew words about why it was Heisenberg who ini-
gumentsfor the inevitability of acausality and the indispensabi
lity of tially led the campaign against causality are in place here. The interpre
classical concepts is the main focusof this chapter.
tive attempts of quantum theorists took place against a hrasicepoume oO
Acausality and the Indispensability of Classical Conce
philosophical controversies about the changed status of space an time
pts concepts implied byEinstein’s relativity. Young Heisenberg witnesse
As I have argued in previous chapters, the stand agains the emotional and politically charged confrontation between Einstein
t causali ty, espe- and the neo-Kantians (Cassidy 1992, 96-98). To ask whether the ver-
cially in Heisenberg’s able hands,crystallized as a responseto
the threat dict passed on the a priori status of Newtonian-Euclidean space and
posed by Schrédinger’s competing theory. Schrédinger’s declar
edambi- time did not hangoverthe a priori status of causality as well was natu-
tion wasto eliminate the “monstrous”—as he called them—
quantum ral. An intensely ambitious.man, whoaspired, following Einstein’sox
jumps, replacing the G6ttingen-Copenhagen research progr
am with a ample, to leave his mark on the philosophy of nature and the pul ic
continuous, causal alternative. It was in response and in opposi
tion to imagination, and not onlyto solvescientific “puzzles, must have kep
the threat from Schrodinger that Heisenberg resurrected
the founda- an eye on this issue. None other than Heisenberg’s teacher, Arnold
tional status of discontinuity and swiftly developed argume
ntsagainst Sommerfeld, posed the high-stakes challenge. Discussing the pea
causality.
: state of physics” andreferring to the causality issue just a few Ont s
As Ihave argued, the probabilistic formalism of quant
um mechanics before Heisenberg’s uncertainty paper, Sommerfeld called for a “new
was developed by Dirac and Jordan as a direct extens
ion of Heisen- Kant” (1927, 235). Heisenberg wanted to be the new Kant—in his ini-
berg’s fluctuation paper (1926a), in which, contra Schré
dinger andhis tial presentations of the uncertainty principle to academic saunas,
supporters, Heisenberg claimed that for two atoms in
resonance, ex- he always described the abandonment of the “Kantian category o cau-
change of energy must occurvia discrete jumps. Onecannot
obtain the sality” as a natural continuation of Einstein’s overthrow of soe
energy of an atom as a functionof time, argued Heisen
berg, but only space and timeas formsofintuition(see, for example, Heisenberg 1928,
the time averageof energy, or probability that the
energy takes a par-
ticular value. Similarly, for any dynamic variable,
both Dirac and Jor- Ta the philosophical framework of the orthodox interpretation
dan postulated, one can only calculate the probab
ility that it lies arose initially as a forced response to the opposition and developed
 196 Chapter Nine
The Copenhagen Dogma 197
subsequently into a huge project for the legitimation of the Gottingen-
istic lawfulness prevails (on oneside, because of the laws of classical
Copenhagen program.? Because an argument of “finality,”
or inevi- physics, and on the other side, because of the differential equations of
tability, is perhaps the strongest way to advance any sort of claim,
the quantum mechanics). If we supplementthe variables at the dividing
thetoric of finality and inevitability permeates many philosophical
ar- line by some “hidden variables” and then “move”the cut to a new
guments in the Copenhagen framework. Following Bohr, Heisen
berg place, we will come into conflict with the causal lawsthat prevailed at
presented the search for a causal quantum theory, not
merely as hope- this place to begin with (the situation becomes overdetermined).
less, but as downright meaningless. Heisenberg argued that “there
This argument is, of course, unconvincing. If we move the cut into
seems to be the strongest evidence”for the “final” renunciation of
cau- what wasthe classical domain, it does not necessarily mean that the
otis iit OS,

sality and objectivity. The concept of “causality,” after the

discovery of situation becomes overdetermined—the hidden variables for the cut
“quantum mechanics became as devoid of meaning as the concep
t of can turn out to be the existing variables in the macrodomain, as when
the “end of the world” after the discoveries of Columb
us and Ma gellan Bohm supplemented the quantum description with position variables.
(Heisenberg 1934, 17). Bohr, claiming that acausality was
“unavoid- It is not too hard to find serious flaws in other arguments advanced
. able,” ridiculed the “odd wish” for a causal description of quant
um by Heisenberg along similar lines. The more interesting question is:
mechanics: “It appears to meto be exactly that, that we would say that
we are not satisfied with descriptions in English, and we mightlike, when,for what purposes, and in what contexts did Heisenberg use such
because we can’t get theword ‘causality’ .. . to translate it into arguments? a?
San- The crucial part of all “cut” arguments is the need to describe the
skrit, to get everything different” (1937d, 338). And Heisenberg, quot-
ing Bohr, asserted that the structure of Bohm’s “strange” arguments for -measuring apparatus using classical concepts, which is connected with
hiddenvariables is identical to the hopethat “2 X 2 = 5, for this would the morefar-reaching Bohrian claim that to speak of quantum concepts
be of great advantage for ourfinances” (1958, 132). and quantum ontology is meaningless. In manyof his writings, Heisen-
One would expect that the proponentsof the Copenhageninterpre- berg distanced himself from this position. Thushe often said that mea-
tation were in possession of some very strong arguments, if not suring apparatus ‘‘can” (as opposed to the Bohrian “have to”) be de-
for in- scribed by classical concepts. Nor did Heisenberg agree with Bohrthat
evitability, at least for high plausibility. Yet a critical reading reveals
thatall the far-reaching; or one might say far-fetched, claims of inevi- no new concepts and no new language beyondthe classical can be con-
ceived—in fact, while discussing the lessons of the theoryof relativity,
tability are built on shakycircular arguments, on intuitively appeali
ng Heisenberg emphasized that new concepts arise and a new language
but incorrect statements, on metaphoricalallusions to quantum
“insep- develops “which adjusts itself after some time to the new situation”
arability” and “indivisibility,” which have nothing to do with quan-
tum entanglement and nonlocality. Acausality, in the writings
(1958, 175). For example, in “the theory of generalrelativity, the lan-
of Bohr, guage by which we describe the general lawsactually now follows the
Heisenberg, Pauli, and Born,is not a sharply defined concept.?
Because scientific language of the mathematicians” (1958, 177).
acausality is to be used as a tool of legitimation, as a sword to
wield I argued in chapter 8 that mathematical physicists, especially those
against opponents, its meaning changesfrom text to text, from
context who had firsthand familiarity with relativity theory, such as Born,
to context.
Among Heisenberg’s many arguments for the inevitability Fock, and Heisenberg, did not accept the special status Bohr accorded
of acau- to classical concepts. Instead, they held that concepts are justified by
sality, oneof the more frequent and careless onesis the argument based
on the notion of a “cut” between quantum object and classica their place in a coherent mathematical formalism of the physical do-
l measur- main—be it classical or quantum.It is unwarranted, therefore, to give
ing device. This cut is mobile. Heisenberg asserted that thestati
stical special status to classical notions. David Bohm wasofa similar opinion:
aspect of quantum mechanics enters only on the cut, or
dividing line “Mathematics in general andthedifferential calculusin particular have
(1979, 15, 49; 1958, 48—49) because oneitherside of the
line determin- played a key role in guiding the developmentof a clear concept of ac-
2. On the “Copenhagen hegemony,” see Cushing (1994b),
and on the “diffusion of the
celerating motion. . . . it is practically impossible to gain such a clear
Copenhagenspirit,” Heilbron (1987). concept on the basis of everyday experience alone, or even on the basis
3. In this chapter, I will follow Bohr, Heisenberg, and of the laboratory experience alone.” Similarly, our concepts of waves
Bohm in using “determinism”
and “causality” as interchangeable notions, despite
the differences between them. These
differences are not important forthe issues discussed here. come not from watching water waves, but from the mathematicaltreat-
ment ofinterference and wave propagation (Bohm 1957, 97-98). Bohr
 198 Chapter Nine
The Copenhagen Dogma 199
was unique among quantum physicists in his denial of
the possibility indispensability of classical concepts (and Heisenberg’s inability to con-
of new concepts. One cannot avoid the question, then: why
and when ceive of categories other than wavesandparticles) that lead him to a
did Heisenberg, very much a mathematical physicist, suppor
t this pe- belief in finality, my claim is rather the opposite. It was his vested in-
culiar doctrine of Bohr’s?
terest in finality that led Heisenberg to insist on the indispensability of
In the cut argument, analyzed above, Heisenberg described
the mo- classital concepts in a local and opportunistic fashion. There was nei-
bility of the cut as “particularly important.” The next senten
ce reveals ther belief nor commitment on Heisenberg’s part—only a selective and
whyit is so important: “An appreciation of this fact also helpst
o dis- opportunistic use of Bohrian doctrine in thosecircumstances where
pose of an objection frequently made against the finalit
y of quantum Heisenberg’s aim was to argue against the oppositionand for the fi-
mechanics” (1934, 16, my italics). Heisenberg immediately
proceeded nality of the Copenhagen orthodoxy. This isa characteristic example of
to the “more general question of the finality” of changes promo
ted by a powerfulsocial strategyof legitimation disguised as an abstract theo-
Copenhagenproponents and concludedthat the renunciation of objec-
tivity and causality would befinal (1934, 17). Similarly, retical argument. eB
Heisenberg pre- It is quite instructive to see how Heisenberg argued in those articles
sented the flawed cut argumentin anotherlecture a year
later, again where he discussed the lessons both ofrelativitytheory and of quan-
using it as an argument for “thefinalityof the statistical character of tum physics. For the mainlesson ofrelativity, as Heisenberg presented
quantum mechanics,” and against “any hope” of “completing quantum “it, was the invalidity of the a priori status of classical concepts, while
mechanics on a determinist basis” (1935, 48). Again, in anothera
rticle, the defense of the Copenhageninterpretation forced him to claim that
Heisenberg claimed thatthe “root of the statistical character of quan- classical concepts are indispensable and final in an‘a priori Kantian
tum theory”is the “tension” between the need to use the concepts of sense. Ingenious Heisenberg had no problem avoiding the impasse: he
classical physics and “the knowledgethat these concepts do notfit ac-
invented the notions of “practical a priori” and “relative synthetic a
curately.”” Heisenberg mentioned that “it has been suggestedthat one priori,” whichare, in fact, contradictions in terms. The only function of
should depart from classical concepts” and search for radically new these self-contradictory notions, as they appearin Heisenberg’s works,
ones that “might possibly lead back to a non-statistical, completely ob- is to argue for the a prioristatusof classical concepts in the context of a
jective description of nature.” He then immediately denied such a pos- defense of the Copenhagen framework,against Heisenberg’s own bet-
sibility, repeating the Bohrian argumentthat “the concepts ofclassical ter judgment whendiscussing relativity (1958, 90-92). Bohm, after his
physics-are just a refinementof the concepts of daily life” (1958, 56).
_“conservasion” had taken place, understood that arguments for the in-
Heisenberg’s denial was a forceful one: “We cannot and should not re- evitability of acausality cannot be compelling. Theyare, in fact, circular:
place these[classical] concepts by any others. . .. we cannot and should “The conclusion that there is no deeper level of causally determined
not try to improve them”(1958, 44).4 motionis just a piece of circular reasoning, since it will follow only if
I argued in chapter 8 that the Copenhagen interpretation cannot be we assume beforehandthat no such level exists” (Bohm 1957, 95).
reconstructed as a coherent philosophical framework, for it is more
adequately described as a collection ‘of local, often contradictor
y, ar- Operationalism: From Consistency to Inevitability Arguments
guments embeddedin changing theoretical and sociopolitical circum
-
stances. Heisenberg’s contradictory statements about the necess
ity of Bohr’s arguments for the inevitability of acausality and complementar-
classical concepts in different contexts is more evidence for
my claim. ity similarly cannot withstand close scrutiny. I have shown that Bohr’s
From this discussion it becomesclear that Heisenberg insisted on
the arguments for inevitability are in fact arguments of consistency in dis-
indispensability of classical concepts in those contexts,
and, as far as I guise (Beller 1993). “The uncontrollable disturbance,” or uncontrollable
am able to determine, only those contexts, where he attempted to argue
for the finality of the indeterministic Copenhagen version of quant interaction between the measuring and the measured, is the crucial
um componentof all of these arguments and the basic support on which
mechanics.
Returning to Bohm’s contention that it was Heisenberg’s belief in Bohr’s web of arguments for complementarity rests. Bohr employed
the different strategies to make this uncontrollable interaction, founded on
4. David Finkelstein’s pointed comment ontheselines strikes the
heart of the matter: the “indivisibility of the quantum ofaction,” a fundamental conceptin
“This seemslike too muchprotest; if we cannotthen the should-n
otis surely otiose, and quantum philosophy.
if we should not then we probably can” (1987, 291).
Yet how exactly does the “uncontrollability of the interaction” follow
 200 Chapter Nine
The CopenhagenDogma 201
from the “indivisibility of the quantum of action,” and what do these hocness of many similar arguments. One striking example is Heisen-
terms mean? Their use and their interconnection differed, in fact,
sig- berg’s reversalof the idea of disturbance—ifthe original idea of distur-
nificantly before and after 1935 (the EPR challenge). Before 1935, Bohr’s bance was based on the indivisibility, or finitude, of microscopic par-
arguments were incorrect and were based onthelater discredited idea
ticles, now the “unavoidable disturbance”is seen as the reasonfor their
of disturbance. After 1935, the arguments were tautological or
circular: indi¥isibility- “Nature thus escapes accurate determination . . . by an
textual analysis reveals a web of terms, where “indivisibility of the unavoidable disturbance which is part of every observation. ... In
quantum of action,” “individuality of quantum phenomena,” “unsur-
atomic physics it is impossible to neglect the changes produced on the
veyability of interaction,” and “uncontrollability of interaction” have observedobject by observation.. .. The supposition thatelectrons, pro-
the same meaning and are interchangeable. tons and neutrons, according to modern physicsthebasic properties of
_ The following lines are typical of Bohr’s pre-1935 writings: “Quan- matter,arereally the final, indivisible particles of matter, is only justi-
tum of action means thatit is not possible to distinguish between the
fied bythis fact’ (1979, 73).5
. phenomena and observation. . . . we cannot make the interaction be- Heisenberg’s and Bohr’s arguments for the foundational, a priori
tween the measuring instrument and the object as small as welike. status of uncontrollable interactionare illegitimate. The uncontrollabil-
Now,that means that phenomenaareinfluenced by observation’ (1930, ity of interaction has one, and only one source—it follows from the
134). Before 1935 the “indivisibility of the quantum of action” simply mathematical framework of the quantum theory.It is necessary to en-
‘meant“finitude” of the quantum of action. Measurement “disturbs” sure the consistency of quantum theory whenthepossibilities of mea-
phenomenaprecisely because of the atomic constitution of measuring
bo3;
ey
devices. The underlying intuition was that becausein the quantum do-
surement are taken into account. Bohm, after abandoning the ortho-
dox Copenhagen stand, understood this clearly. Thus, discussing the
main phenomena and measurementinteraction are “of the same order
uncontrollability of the interaction in Heisenberg’s y-ray experiment,
rege neeMesMRCNG
EB

j
ey ae

fe
of magnitude,” such interactions, unlike in the classical case, cannot
be Bohm pointed out: “The disturbance is unpredictable and uncontrol-
neglected or accounted for (chapter 7). This intuitively appealing argu- lable becausefrom the existing quantumtheory there is no way of knowing
mentis, of course, incorrect. The concept of disturbance is an inconsis
or controlling beforehand what will be the precise angle, with which
oe

-
tent one:it presupposes the existence of objective exact values that are
the light quantum will be scattered into the lens” (myitalics). Bohm. /
changed by measurement, contrary to the desired conclusion of their realized that the indeterminacyrelations are “primarily ... adeduction
indeterminacy. Moreover, the finitude of interaction does not assure
from quantum theory in its current form,” and that attempts to prove
uncontrollability—many classical measurements involve finite inter- their “absolute and final validity” are based on nothingbut ana priori,
actions that cannot be neglected yet can becalculated. illegitimate insistence on thefinality of quantum theory (1957, 82 n).
After 1935, when the EPR challenge underminedthe notion of distur- Thelegitimate function of uncontrollability argumentsis to argue the
bance, Bohr’s reasoning did an about-face. The uncontrollability of in- consistency, rather than the inevitability, of uncertainty and comple-
teraction was then tied inseparably to the necessity of describing mea- mentarity. Bohr (1935a) did discuss certain canonical thought experi-
suring devices using classical terms only, neglecting “in principle” their ments along these morelegitimate lines. He argued, for example, the
atomic constitution.
impossibility of measuring the position of a particle in a setup measur-
In Bohr's laterwritings the uncontrollability of the interaction follows ing its momentum. Using a movable diaphragm, we can measure a
from its “unsurveyability,” which in turn follows from the dogmati-
particle’s momentum (by using an appropriate test body) before and
‘cally postulated essential difference between quantum object and
clas- after the passage of a particle throughtheslit. By applying the Jaw of
sical measuring apparatus.If, in Bohr’s pre-1935 writings, uncontro
l- conservation of momentum to a system consisting of a particle and a
lability necessitated acausality, in his later writings it is the statistical
moving diaphragm wecalculate the momentum ofa particle passing
character ofquantum theory that leads to the inseparability of the
mea- throughtheslit. However, because we have exact knowledgeof the mo-
suring and the measured and to the uncontrollability and
unsurvey- mentum of the diaphragm during, or immediately after, the particle’s
ability of the measuring interaction.
This reversal of whatis to be explained and whatis foundational,
of 5. In writings by Bohr and Heisenberg, even whentheoriginalideaof disturbance is
explanans and explanandum, is frequent in Bohr’s and Heisenbe
rg’s writ- explicitly rejected, its imagery is used implicitly in metaphorical, intuitively appealing
ings. This reversal is an eloquent illustration of the circularity
and ad allusions (see chapter 12).
 202 Chapter Nine
The CopenhagenDogma 203
passage, we block the possibility of also knowing the positio
n of the This mutual exclusivity, aided by operationalism, would soon develop
diaphragm,in accord with the uncertaintyrelations. Consequently __
, we into arguments for the inevitability of uncertainty and complemen-
are denied knowledgeof the particle’s position, which is identical
with tarity. The connection between operationalism and inevitability was
that of the diaphragm (chapter 7; see also Beller and Fine 1994).
Why substantially strengthened in Bohr’s response to EPR. As Arthur Fine
doesthe diaphragm obeythe uncertainty relations? Either we juststat
eit and Yhave argued (Beller and Fine 1994), Bohr wasonly able to counter
(it applies to quantum objects), and then we deduce uncertainty
for the the challenge of EPR by using an operational and relational definition
particle (this is, of course, an argument of consistency), or we
say that of concepts. After EPR, operationalism-becamethe strongest tool in
while measuring the momentum of the diaphragm with a test
body, legitimationargumentsfor the inevitability of quantum theory.* It is in
the diaphragm undergoes uncontrollable interaction. We deduce
con- "thiscontext that Bohr often used the word “logical’”—the inevitability of
sequently the uncontrollable interaction between the particle and
the - complementarity and uncertainty follows; in his words, from “simple
diaphragm— again,these are elementaryconsiderations of consist
ency. logical analysis,” and it leads. to “the situation where the very concept
The crucial point of these thought experiments is that momentum
of determinism losesits possibility of logical application” (1957b, 601).
and position cannot be measured simultaneously. This is, however, not
With the “dictatorial help” of operationalism, to use Schrédinger’s
satisfactory to Bohr. Whyisit notsatisfactory? The answeris clear from
(1935, 157) apt expression, the central Copenhagen dogmas became un-
Bohr’s own words. He warnsagainst “misunderstanding by expressing
assailable andfinal. ; ‘
the content of uncertainty relations . . . by a statement as ‘the positio
n — Now wecan return to Bohm’s claim thatit was thepositivist commit-
and momenta cannot simultaneously be measured with arbitrary accu-
racy.’ . . . such a formulation . .. would notpreclude the possibility of ~- "ments of the foundingfathersof the Copenhageninterpretation that led
a future theory taking bothattributes into account” (1937c, 245). What them to believe in the finality of quantum theory and the impossibility
is the strategy of precluding a theory, such as Bohm’s, whereparticles of revising it along more objective or deterministic lines. As in the case
of the inevitability of classical concepts, the situation is the opposite:it
have well-defined position and momentum, whichhowever cannot be
wasthe need, or the desire, to argueforfinality against threats from the
measured simultaneously according to the uncertaintyrelations? How
do we argue that Bohm’s theory is impossible in principle? How do we opposition that led Heisenberg and Bohrto take a forceful operational
transform consistency argumentsinto those for inevitability? stance. | argued in chapter 8 that the proponents of the Copenhagen
With the benefit of hindsight the answeris obvious. Simplypostulate interpretation did not have, collectively or even individually, a consis-
that what cannot be measured—does not exist. By defining concepts tent positionon the realism-antirealism issue and that they employed
different, often contradictory arguments, depending on the local con-
operationally through a procedure fortheir measurement and then ap-
plying the quantum formalism to an analysis of the measurement pro- text. We have nowidentified the context in whichpositivist operational
cedure, we will obtain nothing but deductions from the quantum for-
arguments were selectively employed. Positivism in quantum philoso-
malism (such as, for example, the uncertainty relations). In this way an phy has two central strands — operationalism and instrumentalism.In-
illusion is created that features of the theory (such as uncertainty) be- strumentalism implies that the quantum formalism is merely a tool, an
long to the very definition of the concepts used and that they follow algorithm for the description and prediction of measurement results.

$fe
inevitably from logical analysis of the conditions of experience. The instrumentalist strand, denying the possibility of a quantum on-
Thought experiments were indeed initially used by Heisenberg and tology, while emphasizing the Kantian inevitability of the classical on-
tology, is, of course, directly connected with the doctrine of the indis-
Bohrto argue the consistency of the new quantum mechanical scheme.
This was the original thrust of Heisenberg’s analysis in the y-ray " pensability of classical concepts. Both varieties of positivism are aimed
thought experiment. Bohr (1929a), following Heisenberg, argued at arguing the finality of quantum mechanics.
that Were physicists aware that unless they took an extreme and unten-
the uncertainty relations are the basis of the logical consiste
ncy of quan- able positivist stand, what the Copenhageninterpretation offered were,
tum mechanics. Bohr’s considerations of mutually exclusive
experi- at best, considerations of consistency, and notof inevitability? After
mental arrangements wereinitially aimedat proving the consiste
ncy of
quantum theory: the wave andparticle aspects “can never be brought
into direct contradiction with one another, since their closer analysis 6. It is instructive to mention here that Bridgman’soriginal aim of introducing opera-
in tionalism was to prevent the “embarrassing”need for basic revision of physical theories,
mechanical terms demand mutually exclusive arrange
ments” (1933, 5). ensuring their certaintybya strict operational definition of concepts (Beller 1988).
 _204 Chapter Nine
The Copenhagen Dogma 205
abandoningthe orthodox stance, Bohm was awareof this. According
to The project of arriving at true, final, certain, indisputable knowledge
Bohm, Bohr's essential contribution was to show that one can “con-
sistently . . . renounce the notion of unique and precisely defined was, of course, not a new one. Classical physics, despite the empiri-
con- cist revolt against scholasticism, inherited the Aristotelian idealof sci-
ceptual models in favor of that of complementary pairs of imprecisely
ence.as demonstration, of arriving at certain conclusions from certain
defined models” (myitalics). What Bohr was able to demonst
rate was premises. The “father of empiricism” himself, Galileo Galilei, used such
that such a use of complementary concepts provides “a possible
way of scholastic demonstrative arguments: two bodies of unequal weight must
discussing the behavior of matter in the quantum-mechanical domains”
fall at the samerate, for if they do not(if heavy bodiesfall faster), we
tetne nti

(Bohm 1957,93, myitalics). Not only such dissidents as Bohm,but or-

thodox quantum physicists as well, were aware that manyof the philo- would arriveata logical contradiction.” Heisenberg’s cut argumentsare
sophical pronouncements on complementarity and uncerta peculiarly reminiscent of such pseudological proofs.
inty were It was Kant, of course, who provided the most grandiose philosophi-
“merely consistency arguments. Pauli announcedthat
the central func- cal demonstration of indubitable knowledge—byhis analysis of New-
tionof the notion of complementarity was “to symbolize the
funda- tonian space and time as formsof intuition, and causality asa necessary
“mental limitation of our .. . idea of phenomenaas existing indepen-
dently of the means by which theyare observed. . . . It is indeed this category. Newtonian mechanics received the status of synthetic a priori
limitation that makes the theorylogically consistent” (1945, 99, myital- knowledge—knowledge that, though empirically arrived at, had the
ics). Dirac, who was fond neither of pictures nor of complementarit samecertainty as the analytic knowledge of Euclidean geometry. Bohr's
y, attempts to arrive at true, certain, final knowledge by the mere analysis
conceded thatoné may “extend the meaning of the word‘picture’
to of the conditionsof experience is thoroughly Kantian in spirit. Whatis
include any way of looking at the fundamental laws which maketheir
self-consistency obvious” (quoted in Jammer1974, 13, myitalics). striking, however,is that such Kantian arguments were advanced after
Both Feynmanand Weizsacker realized that Heisenberg’s analysis of a full realization of the bankruptcy of Kantian argumentsfora priori
the y-ray experiment was merely a consistency argument, despite knowledge(the theory of relativity being thefinal stroke), andby those
the physicists who, excepting Bohr himself, had a thorough knowledge of
fact thatit is often presented as an argumentfor inevitability. Feynma
n relativity theory. Only “extrascientific’” reasons can plausibly explain
announcedin his Lectures on Physics: “It was suggested by Heisenb
erg this fact. Not surprisingly, then, what the orthodox physicistsadvanced __
that the new laws of nature could only be consistent if there were
some as argumentsfor inevitability were, in the best case, arguments of con-
basic limitation on our experimental capabilities not previously
rec- sistency. It is enlightening to mention here the views of the Géttingen
ognized” (Feynman, Leighton, and Sands 1969, 1-9, myitalics).
And, mathematician David Hilbert, with whom the mathematical quan-
more eloquently, Weizsacker: “The impossibility of simultaneously
ob- tum physicists were in close contact. Pitowsky outlines Hilbert’s stand
serving precise values of position and momentum does not
mean the succinctly: “Consistency, or rather the avoidance of a paradox,is ulti-
pseudo-positivistic nonsense ‘what cannot be observed does not exist,’
butthe test of consistency: ‘what does not exist in the theory cannot be mately the only safeguard in the pursuit of knowledge. We would have
observed in an experiment describable by the same theory’ ”(1987, 278). liked to do better for sure, but we haveverylittle choice in the matter”
In this argument Weizsacker was following his teacher Heisenb (1994, 119). -
erg, The Copenhagen interpretation, andits rhetoric ofinevitability, rests
whorevealed that he regardedall of Bohr’s complementarity
elabora- on twocentral pillars—positivism and the doctrine of the necessity of
tions as demonstrating the consistency of the quantum formalism
by classical concepts. Both are unnatural for working mathematical physi-
an analysis of the possibilities of measurement. Heisenberg
considered cists. No wonderthe Géttingen-Copenhagen physicists, Heisenberg in-
all these arguments “trivial”: “Many experiments were
discussed and cluded, often denied their adherence to positivism. No wondercontra-
Bohr again successfully used the two pictures, wave andpart
icle, in the dictory statements abound concerning the doctrineof the necessity of
analysis. The results confirmed the validity of the relation
s of uncer- classical concepts.
tainty, but in a way this outcome could be considered
trivial, because
if the process of observation isitself subject to the
laws of quantum
theory, it must be possible to representits
results in the mathematical 7. If we combinethe two bodies of unequal weight, the composite body should fall
scheme ofthis theory” (1977, 5—6). This position undermi slowerthan the heavierof the two would havealone, becausethelighter of the two bodies
nes, of course,
all of Heisenberg’s own argumentsfor inevitability, impedes the fall of the composite body. It also should fall faster than the heavier body,
or finality.
becauseits weightis greater. Hence all bodies, regardless of weight, fall at the samerate.
 206 Chapter Nine
The CopenhagenDogma 207
Bohm on Classical versus Quantum Concepts and on Indeterminism
the prediction and description of experimental results. Bohm called for
Bohr advancedhis doctrineof the indispensability of classic the development of a new language, which would depart from classi-
al concepts, cal intuitions and would be adequate for the new quantum conceptual
on which the whole framework of complementarity stands
or falls, as framework. Bohm also undertook a thoroughcritical analysis of the
an obvious fact. But, as I argued in chapter 8,it was not a doctri
ne that classical idea of the continuity of motionand of classical determinism.
Bohr’s closest collaborators, Heisenberg and Born,
could subscribe to, His aim was to demonstrate that “there is no a priori logical reason for
public support of Bohr notwithstanding. Even those physic
ists eager to their adoption.” The quantum concepts of indivisible transitions and
follow Bohr foundit difficult not to question it. Bohm’
s discussion of incomplete. determinism are “much more analogous to certain naive
classical concepts and causality, composed before his defect
ion from “concepts that arise... . in common experience”(1951, 144). Not surpris-
the orthodoxy,is instructive.
ingly, Greek philosophers were unable to grasp the idea of continuous
Bohm’s Quantum Theory is rightly considered the best Bohria
ntext- motion, as Zeno’s paradoxes eloquently remind us (1951, 147). In fact,
book. Indeterminism, indivisibility, and wave-particl
e duality—all are ourintuitions revolt against thinking of a motion in which the position
elaborated with rare clarity and force. The arguments are
perfected as and velocity are both precisely determined andare thus in close har-
far as they can be. Because Bohmlabored so hard to make themas monywith the uncertainty principle. Using the example of photogra-
convincing as possible, he could not avoid seeing their weak points.It phy, this point becomes especially clear. Our simplest iatuition of an
is perhaps notincidental that both Bohm and Alfred Landé, another - - object with a definite position is an objéct thatis not moving. To obtain
prominent dissident, defected from the orthodoxy after comple
ting a picture of movement, we mustallow the position to blur slightly.It is
textbooks on quantum theory in the Copenhagen spirit. While the cen- the blurred picture of the speeding car that suggests some space was
tral pillars of the Copenhageninterpretation are elaborated faithfully
in covered during some timethat the car was moving.If we use very fast
Bohm’s textbook, a careful reading finds a powerful undercurrent
of camera and get a sharp picture of the movingcar, we will not deduce
tensions with Bohr—tensions that, whentakento their logical conclu-
motion. Our intuitive presupposition is that “a continuously moving
sions, should have, and historically could have, been the
reason for object has a somewhatindefinite range of positions”—wesimply “can-
Bohm’sfinal break with the orthodoxy. Not surprisingly, these discrep
- not visualize simultaneously a particle having a definite momentum
ancies concern Bohm’s discussion of classical and quantum
concepts. and position.” Bohm concluded that “quantum theory .. . gives a pic-
Bohm treats these issues in chapter 8, revealingly titled “An Attemp
t ture of the process of motionthatis considerably closerto our simplest
to Build a Physical Picture of the Quantum Nature of Matter’:
the concepts than doesclassical theory” (1951, 146).
wording holds a hint to—and possibly a seed of—his future quant
um But does not our concept ofan objectat rest, with fixed position and
ontology.
zero velocity, contradict the uncertainty principle, according to which
Unlike Bohr, Bohm did not arguethatclassical concepts are either in-
an object with a precise position has a completely indefinite momen-
dispensableor natural. In fact, the Bohmian analysis undermines Bohr’s
tum? Bohm arguedthat in fact there is no contradiction. When we think
arguments that classical concepts are extensions of commo
n sense. Nor © of an objectat a definite position, we can imaginethata short time later
did Bohm seem to agree with Bohrthat there can be in principle “no
it will moveto a different position. Which position it will move to we
guantum concepts,” only a quantum mechanical abstract algorit
hm for cannotinfer from the original position—all subsequent positions are
8. Manyphysicists who did not trouble to go into the Bohrian equally consistent with it. This means that any velocity, or any value of
argumen ts in depth had
an easier time following them uncritically. Heisenberg
complained as early as 1930 that momentum, is equally consistent with a picture of a particle in a defi-
physicists had more faith in the Copenhageninterpretation nite position. Put differently, “if we think of an object in a given posi-
than clear understanding of
it (1930, preface).
: tion, we simply cannot think of its velocity at the same time” (1951,
According to Bohm,“The whole development[of a search for
native] started in Princeton around 1950, when I had just finished
a hidden variables alter- 146). Wesee, according to Bohm,thattheclassicalidea of the continuity
my book Quantum of motion is an unnatural, non-commonsensical idea. Ona purelylogi-
Theory. [had written it from whatI regarded [as] Niels Bohr’s
point of view, based on the
principle of complementarity. Indeed I had taught a course on
the quantum theory...
cal basis, there is no reason to choose the concept of a continuoustra-
and hadwritten the book primarily in order to try to obtain jectory in preference to that of a discontinuous trajectory. The idea of
a better understanding of the
whole subject. However, after the work was finished, I looked back
over what I had done “continuous motion developed through our experience with planetary
andstill felt somewhatdissatisfied” (1987, 33).
orbits, projectile trajectories, and the theory ofdifferential calculus was
PonsaaryiORE! Yima Bi
 208 Chapter Nine
The Copenhagen Dogma 209
tailored to deal with such motions.It is only by studying “such
things been used in connection with common experience is that a particular
for a while” and getting used to them that the “succeeding genera
tion force or cause produces a tendency toward an effect, but
beganto take the basic ideas for granted” (1951, 148). This analysi that it does not
s is of guarantee the effect” (1951, 152). Quantum lawfulness, which deter-
course a far cry from Bohr’s assertion of the necessity of classica
l con- mines only statistical trends, is much closer to our commonsense than
cepts because ofthe alleged proximity ofclassical to commonsens
e no- complete determinism (1951, 152).
tions andthe direct accessibility of classical reality to sense percept
ion. Bohmoffereda brief critical analysis of the history ofthe idea of clas-
According to Bohm, there are quantum concepts thatare inheren
tly sical determinism. The Greek idea of “fate’’—the course of events that
different from classical concepts. Quantum energy and momen
tum, de- is beyond the power of humans to change—belonged to one of the
spite being given the same names as: concepts in classical
physics, are earliest conceptions of complete determinism. Yet it was only with the
examples.In classical theory, we regard energy and momen
tum as fun- invention of machinery, which allowed motion to be controlled with
damental properties of matter because they obey conser
vation laws. But precision, that the idea of a mechanistic and deterministic worldview
from a logical point of view they need notbe considered fundam
ental, developed in the sixteenth through nineteenth centuries. When work
because energy and momentum can be expressed as functions of
posi- was done mostly by hand or with the aid of animals, motionscouldnot°
tions and velocities, so they are in a certain sense redund
ant. All the be controlled exactly—one could pushin the right general direction,
classical laws of motion can be expressed in terms of
space-timerela- and push backward if one had gonetoofar. “It is very likely that the
tionships alone. This is not the case in quantum theory. In quant
um modern form ofthe idea of complete determinism wassuggested... by
theory, velocity cannot be defined as’a limit of distance divided by time.
its resemblance to complex and precisely constructed machines, suchas
In quantum theory, momentum, defined by the de Broglie relation,is
to clocks” (1951, 150). Bohm’s analysis of machines—material means of
be considered an independentrather than derived property of matter—
production—as determining the basic conceptual framework of New-
“thestatementthat an electron was observed to have a
given momen-- tonian mechanics resembles the analysis of the Soviet historian of sci- ©”
tum . . . stands on the same footing as the statementthat it had a given
ence Boris Hessen, who in 1931 at a congress in London, launched the
position. Neither statement is subject to further analysis”
(1951, 154). Marxist historiography of science with an analysis of Newtonian me-
Bohm departed from Bohralso in his position on theissue of clas-
chanics along similar lines. It would be rewarding to know whether
sical versus quantum mechanicalreality. He uses the word
“conveni-
ent,” rather than the Copenhagen “necessary,” to differentiate betwee
Bohm wasacquainted with the works of Marxist historians of science ~~ ~
n in general, and with Hessen’s work in particular. Yet the crucial point
classical and quantum processes: “It is convenient to make a
distinc- here is this: Bohm’s analysis of historical rootedness, of the sociohistor-
tion between classically describable processes and those
essentially ical contingency of our most cherished, most uncritically accepted con-
quantum-mechanical in nature.” Such a distinction is convenient when
cepts, does not agree with the rhetoric offinality of the Copenhagen
we deal with large objects, whose features of interest “do notcriti
- interpretation. What Bohm’s analysis rather discloses is that nothing
cally depend on the transfer of [a] few quanta moreor
less.” In direct, more than continuous habit endows concepts with the auraof natural-
though unacknowledged, opposition to Bohr, Bohm asserted: “In
the ness, necessity, inevitability. We need only take a small step to ask
last analysis all processes are, of course, quantum-mechanical
in na- whetherthe samekind ofcritical analysis can be extended, orwill be
ture” (1951, 165).
It is very enlightening to see how Bohm argued against classical extendedin the future, to the Copenhagen quantum philosophyitself.
La- Even the idea of hidden variables surfaces favorablyin this discussion
placian determinism. By presenting classical determinism as a
histori- (1951, 171). Bohm,in his later quantum ontology, was open on theissue
cally contingent rather than necessary notion, Bohm aimedt
o ease the of determinism versus indeterminism. Bohm’s nonlocal ontology can be
way for the acceptance of indeterminism. Yet precise
ly because of its either causalor stochastic. No wonder Bohm wassoonledto his idea of
historical contingency, Bohm implicitly undermined
the Copenhagen understandingnature in terms of an inexhaustible diversity and multi-
stand ofthe inevitability of indeterminism. Bohm argued
for replacing plicity of things, and of understanding scientific theory as approximate,
the notion ofclassical determinism with a more limited
notion of cau- conditional, andrelative truth (1957, 164-65).
sality as determining general tendencies(butnot determi
ning the be- Howstrikingly different from Bohr’s ideasthis trend of thought was!
havior of systems completely). This notion is much closer
to the way we Bohr, whoconsidered dlassical concepts natural and necessary and who
relate to forces or causes in everydaylife: “The idea
most likely to have considered quantum theory “a rational generalization” of classical
ta
210 Chapter Nine

theory, could believe (as opposed, perhaps, to the more opportunistic

Heisenberg) that his analysis of quantum theory was really the only CHAPTER 10
possible one. Bohr was impatient with those who questioned him: ey
“How could philosophers not understand, that this was an objective
description, and the only possible objective description?” (interview
, Constructing the Orthodox Narrative
with Bohr, AHQP; Bohr to Born, 1953, AHQP). Astheofficial spokes-
manof the orthodox interpretation, who hadto fight against opposition
The tendency in many historians [is] to write on the side of Protestants and
such as Bohm’s, Bohr preachedclosure, finality, inevitability, and the Whigs, to praise revolutions provided they have been successful, to emphasize
impossibility of alternatives. Howfitting for Bohm, a victim ofthis ap-_ certain principles.ofprogress in the past and to produce a story which is
proach, to develop a diametrically opposed view of science—pleading the ratification if not glorification of the present.
for tolerance, for creative plurality, for peaceful theoretical coexistence,
for a free play of imagination, for friendly, open-minded, and joyful _ Herbert Butterfield 1931
scientific cooperation and communication (Bohm and Peat 1987).

Introduction: “Whiggish” History and “Winner's” Strategies

How does one construct the orthodox narrative, in which the devel-
opment of “disembodied” ideas seems to follow so persuasively, if
not inevitably? How does one retrospectively construct coherent con-
ceptual frameworks from a multitude of contradictory attempts, pre-
dispositions, vacillations? How does one suppress the polyphony and
temporality of the creative process bytimeless, static paradigms? How
does one obtain what professional historians, not without disdain,call
a “Whiggish”history?!
It is often awkward to deduce the past from the present, in life as in
science. The following conclusion of Schrédinger’s biographer sounds
strange: “Considering what a good swimmerhelater became, he prob-
ably could swim well enough at ten” (Moore 1989, 19). Yet there is also
something very natural, almost inescapable, about Whiggishness. We
often assign meaning retrospectively, after some accumulation of infor-
mation that allows a perspective to take shape. Weall, so it seems, pro-
ject our present knowledgeof the outcomeof ouracts onto an appraisal
of our owndecisionsin the past (Fischhoff 1975, 288). Writing history
backward need not be a cynical, self-serving act, though it sometimes
is. We have a whole spectrum ofpossibilities—from innocentattempts
to understand the present by looking for coherence and meaningin the
past, to conscious manipulations of history. The past is often manipu-
1. The passage from Butterfield quoted in the epigraph describes narratives of British
political history. Historians of science have adopted the expression “Whiginterpretation
of history of science,” or “Whiggish history of science,” to characterize those works that
interpretthe past in termsof the present, ignoring historical material that cannotbe inter-
preted as having contributedto the “triumph”of the present.
212 Chapter Te“
(ions!ru (ling 9hr ( )rthndni \

”twine ‘
lated in Whiggish narratives, which make the “Winners" 1 mathematical formalism as havin
00k ,,right"
.n,
so naturally.
Not all scientists are aware, as Feynman WaS, tha t What h
the matter—of—fact history of their discipline is merely "a 80 t e ta e as :1:fighilosophy.3: fioegfileg‘gmarity: ”Thegpl::fp::r‘;:‘::hinamaimm
for a ComprehenSl P n on cuff-1plememaW.hm MM
the so-called quantum meChanical formali; .. '3 av E.‘ b???) CT\'<
in

tionalized myth—story.” 2 The history of quantum me Chanjrct

Of COnVEn-
written from the "winner’s” perspective, With heavy relianc S ls usual]V rhetoric
As in the case of the
wife 23:?
of inevitability, 19.
ollections of Bohr, Heisenberg, and such ardent followerS
This Whiggish, winner-oriented perspective is Clearly displa Osenfeld.
ample, in the widely used collection Quantum TheoTy and Measu Yed,f ex.
01‘
a:
the reé. {in explanans Often
we clailn
berg’s
change
eventsabout the
1, } :
are interc anged
the Orthodox
aces, so

earllly réahzatlonof'the
In
narratix-:ndum and
freely All need example 13:?2min“
QXtI‘Eme
to introduce
and

ONheeler and Zurek 1983). The volume h

begins 'With Bohr’8
- Papers remem
account
server into quantum mec
anical
description, which supposedly
£3”
the past (Bohr 1949), and only later t e actual historical of BOhIIS theory Of the atom.
There wa?indepe tilt:
reason0 to disbelieve t
800d ndent
'
the course of events was ob]ective and .

ing those of the Bohr—Einstein debate, are introduced. Bohr's

1:cluo. Bohr: 31:25:95;
structure" (1979, 14). No less far-fetched are BOm's Presentations$01138
consequences of these discoveries have led to
h]:
HeisenbeI: Pet
prefaced by six pages of ”commentaries” by Bohr and 18 of
less than half a page left to a “commentary” by Einstein, Only
erate, concentrated effort allows a reader to avoid approaching the i 53:1th- realized theclaimed.
tory. Born
“inev1table (incorrectly, as I Wlll shortly argUe) that Bohr funv
indeterminisrn of quantum transitions fth
the beginning—‘3 realization that led immediately to a blurring of the
pers in this volume from the Copenhagen perspective. pa-
In this chapter and the next I will describe the construction of histori- separation between observer and Observed}
cal narratives by which naturalness, or even inevitability, is ascribed by The/line between such manipulation of history and deliberate de~
the orthodox to such central notions of the Copenhagen interpretation Ception is very fragile. Heisenberg often stated in his recollections that
as quantum jumps, the impossibility of causal space—time models, Schrodinger could not obtain even Planck’s law without the quantum
wave-particle duality, Born’s probabilistic interpretation, the indispen- postulate—yet Schrodinger, of course, did, and published a paper at
sability of classical concepts, and complementarity. Orthodox physi— the time with a derivation of Planck’s law (Schrodinger 1927b). The
cists constructed the past as a major reinforcement of their rhetoric of Gottingen-Copenhagen physicists often referred to John von Neumann’s
inevitability. Different strategies—from describing later insights as the ”impossibility” proof as a conclusive strike against hidden variables.
Yet in 1935 Grete Hermann, Heisenberg’s student, published an essay
outcomes of earlier efforts, to outright deception—are employed. The in which she argued that von Neumann’s proof is faulty—no other than
complex, multidimensional theoretical activity of the past is conflated Heisenberg wrote a favorable preface to her essay.S Pauli also real—
to a single perspective—the orthodox one. _ . ized, it seems, that von Neumann’s proof was deficient; he argued in
Bohr proclaimed his research program—the rational generallzamn his lectures on wave mechanics that ”no proof of the impossibility of
of classical physics incorporating the quantum of action—to be an un- extending [completing quantum theory by hidden variables] has been
questionable task shared by all physicists.3 An extreme, yet
istic case is Bohr’s discussion of “Planck’s discovery of the characteri
univeféas
given” (1973, 12). Nevertheless, Heisenberg and other orthodox physx-
cists continued to use von Neumann’s “proof” as an irrefutable argu-
quantum of action” leading directly to the realization of “regulariti- Irient for indeterminism and as a potent weapon against those who con-
of a novel kin ," whose elucidation brings us directly to the
insfegb Sldered a deterministic version of quantum theory possible. ,, .
rability between the ”objects under investigation and our tOOIS the
servation” and to complementarity. Bohr subsequently presente 0(i The magic force of the words “proof” and “von Neumann 15 de-
scrled vividly in Paul Feyerabend’s autobiography: ”He [Bohr].came
a PUblic lecture. . . . At the end of the lecture he left, and the discussmn
for
1’
which
. I
l5
4

2»- , W13“ have just outlined is what I call a ’physicist’s history Of

_ What I am telling YOU is a sort of conventionalized myt
P1113513}!
‘5 i
that the
not neC’ 30hr ' - Stressed, from the very beginning, the new features
'

(:1th339$:32mm:
mel

their students, and those students tell to their students! an W” (Feyn' termjxiistic character of the transitions, the appearance of _
arsobserved
and the
,

the mual historical development which I do not really kno thing .This means the end of the sharp separation of the obiec
less
‘ (Born 1953a, 126). Neumann’s proof, are d15'
616‘s discovery confronted thSiCists wai-zilazion 0 9 ‘t, as well as other criticisms of von
'

(1974, 253—339).
,

dream of action in a rational genera 1

 214 Chapter Ten
Constructing the Orthodox Narrative 215
proceeded without him. Some speakers attacked his qualitative argu- beginning. One ofthefirst objections, that of clairvoyance, was raised
ments—there seemed to be lots of loopholes. The Bohrians did not by Ernest Rutherford ina letter to Bohr: “There appears to me one
clarify the arguments; they mentioned the alleged proof by von Neu- grave
difficulty in your hypothesis. . .. how does an electron decide with what
mann, and thatsettled the matter. Now I very much doubtthat those who frequency it is going to vibrate and whenit passes from one stationary
mentioned the proof, with the possible exception ofone or twoof them, state to another? It seems to me that you would haveto assumethat the
could haveexplainedit. lam also sure that their opponents hadno idea electron knows beforehand where it is going to stop” (quoted in Pais
ofits details. Yet, like magic, the mere name ‘von Neumann’andthe mere
1986, 212).
word ‘proof’silenced the objectors” (1995, 78). Another unsettling problem was that the observed width of spectral
enh itM

lines wastotally incompatible with the conceptof instantaneoustran-

Discontinuities and Quantum Jumps sition. Sharp spectrallines (small Av) necessitated a wavetrain consist-
ing of at least 10° waves, and therefore a finite (noninstantaneous) du-
Discreteness is often perceived as the mostbasic feature of the quantum ration for the emission of all the waves (Slater 1975b). As Schrédinger
paradigm. This wasso at least in the pre-Bell era, when the orthodox later putit, the radiating of such a wavetrain “would use upjustthe
argument wasthat both indeterminism and antirealism follow
from the - average interval between two transitions, leaving the atom no timeto
indivisibility of the quantum of action; in the post-Bell era entangle- be in thosestationary states, the only ones of whichthe theory gave
mentis considered to be the most characteristic nonclassical feature
of. a description” (1952a, 113). Though physicists could not describe the-
‘the quantum world. Yet the disagreement concerning discontinuity be- mechanismofelectrontransitions, contrary to whatis usually assumed,
. tween the orthodox and the opposition was not aboutthe presence of © they did regard the question as worthy of consideration from the begin- .
discontinuous phenomena in the quantum domain, butrather about ning. Bohr himself did not exclude the possibility that eventually the
the irreducible, a priori status of discontinuities. Because the.a priori mechanism of transitions would be understood (Bohr 1918). ,
status of discontinuities cannot be conclusively proved, despite the or- _ Retrospectively, two features became inextricably linked with the no-
thodoxrhetoric of its inevitability, it should come as no surprise that tion of quantum jumps: indeterminism and the impossibility of space-
the orthodox commitmentto the irreducible status of discontinuity, and. time visualization, or the “indivisibility” of quantum. transitions. Yet
belief in the real existence of quantum jumps, crystallized as the end these connections were, as I have mentioned, by no meansclear from:
product of a complex historical process. As we saw in Heisenberg’s the beginning. Discussing the notion of quantum jumpsand the proba-
case, both theoretical considerations and psychosocial factors playeda bilities of transitions in 1925, Pauli admitted that the time was not yet
crucial role in this process. A strong belief in, or strict commitment to, ripe to come to any definite conclusion on the issue of determinism
any metaphysical presupposition acts too muchlike a straitjacket in a (chapter 4). Claimsin Bohr’s later writings contrast sharply with the
creative, conceptually fluid phase of scientific activity. Thusit is the historical situation. Describing in 1955 the early developmentof atomic
existence of commitment, rather than the lack ofit, that needs explana- theory, Bohr pronounced: “The point of departure became herethe so-
tion. Often “commitment” develops to precisely those concepts most called quantum postulate, according to which every changein the en- ~
vigorously challenged by the opposition (see chapter 13). Discontinui- ergy of an atom is the result of a complete transition between two of
tes and quantum jumps becamethe loci of the Copenhagen paradigm
its stationary states. . . . It was evident that no explanation of the indi- -
because they were strongly challenged by the opposition from the ad- visibility of the transition processes, or their appearance under given
vent of the new quantum theory. conditions, could be given within the framework of deterministic de-
The concept of discrete stationary states of an atom and instantane-
scription” (1955a, 86-87, myitalics). Similarly, discussing the causality
oustransitions between them was Bohr’s mostfruitful, ingenious,
and problem, Bohr argued: “These so-called quantum postulates... impl[y]
blatantly nonclassical contribution to physics (Bohr 1913).
This idea en- an explicit renunciation of any causal description of such atomic pro-
abled Bohrto lay the foundations of quantum physics, by connecti
ng cesses. In particular, as regards its possible transitions from a given
atomic physics with Planck’s hypothesis, solving thecritical problem
of
the stability of atoms and, unexpectedly, deciphering the whole 6. In any case, questions regarding the actual nature oftransitions—whether they are
mys-
tery of the complex spectral “music” to boot. This incredibl instantaneous or of short duration, whether they would or would not play anyrole
in
e concep- optical theory (Bohr, Kramers, and Slater 1924; Kramers and Heisenberg 1925; Bridgman
tion, kindling both admiration and disbelief, was problematic
from the 1927)—were important and meaningful.
eine
 216 Chapter Ten
Constructing the Orthodox Narrative 217
stationary state to other stationary state, . . . the atom maybesaid to be bilistic framework of transformation theory, into Born’s probabilistic
confronted with a choice for which . . . there is no determining circum-
interpretation, and into Bohr’s original formulation of complemen-
stance” (1939, 385). tarity. After 1927, it became a foundational notion in the Copenhagen
This retrospective rewriting of history took place shortly after Bohr, interpretation. No wonder the attempts to challenge the irreducibility
Heisenberg, Pauli, and Born closed ranksat the congress at Como and and instantaneity of quantum transitions provoked an intense counter-
the Solvay conference in Brussels in 1927. What was in 1925 only an attack by the Gottingen-Copenhagen camp. Andbecausethe orthodox
“unsolved issue” (Pauli 1926a, 11-12), Bohr in 1929 presented as
a clear- strategy wasto presenta free theoretical choice as an inevitable concep-
cut, uncontroversial, “conscious” theoretical choice in favor
of indeter- tual development, we should notbe surprised at the inconclusive, high-
minism: “Only by conscious resignation of our usual demands. for vi- pitchedrhetoric of theorthodox and the opposition alike. The 1952 con-
sualization and causality was it possible to make Planck’s discovery
troversy between Born and Schrédinger about quantum jumps, which
truitful in explaining the properties of the elements” (1929b, 108). Later
occurred in the pages of the British Journal of the Philosophy of Science,is
Bohrpresented his overthrow of causality as self-evident: “Since, how-
a goodcasein point.
ever, no description of a state of the physical system .. . can determine
Throughouthislife, Schrédinger sought to discredit Bohr’s concept
the choice between different individual processesof transition to other of quantum jumping. His forceful attacks culminated in the 1952 paper
states, we have obviously to do with a situation which lies beyond the “Are There Quantum Jumps?”In this paper he also presented some
scopeofthe classical idea of causality” (1950, 512, myitalics). If initially objections against Born’s collision treatment and probabilistic interpre-
Bohr was uncertain whether the transition process would be described tation. But, in essence, Schrodinger repeated his old arguments, envel-
in more detail’ (recall also Slater’s objections and their incorporation opedin a good deal of rhetoric. Physical scienceis at present in danger,
into the Bohr-Kramers-Slater theory), he later presented indivisibility, _ he argued, of being severed both from its glorious history and fromits
discontinuity, and acausality as part of the very definition of energetic cultural milieu. Physicists must concern themselves, not only with in-
transitions (instantaneous quantum jumps) in the atom.® : genious theoretical constructs and their agreement with experimental
Yet these energetic transitions need notbe “indivisible”or inherently data, but also with the suitability of their concepts to be absorbedinto
indeterminate. They can be described in terms of.a continuous, causal a general spiritual culture of mankind: “Science is not a soliloquy. It
space-time process if one gives up the Copenhagenparticle ontology gains value onlywithin its cultural milieu, only by having contact with
(as illustrated by Schrédinger’s description of resonance phenomena all those who are now, and whoin future will be, engaged in promoting
or his deduction of Planck’s law, contra Heisenberg), or if one supple- spiritual culture and knowledge” (Schrédinger 1952a, 234).
ments the particle ontology with nonclassical additional assumptions Notall concepts, according to Schrodinger, are destined to survive:
(see, for example, the causal space-time description in Bohm and Hiley “A theoretical science ... where the initiated continue musing to each
1993, chap. 5). The Whiggish history, or the “rational reconstruction,” other in terms that are, at best, understood by a small group of close
is formulated by the orthodox with the primary aim of presenting any fellow travelers, will necessarily be cutoff from the rest of cultural man-
alternative as impossible in principle, and thus of ruling out dissidence. kind,in the long runit is bound to atrophy andossify” (1952a, 109-10).
Ihave argued that the notion of instantaneous quantum jumps was The “abbreviated language” of quantum jumps, Schrédinger argued,is
incorporated into the basic assumptions of the mathematical proba-
not a suitable candidate for long-term survival. There were other “in-
7. Whenever Bohr talked about the role of probability considerations in relation to genious constructs of the human mind that gave an exceedingly accu-
transition processes before the advent of the new quantum theory(oratleast before the rate description of observed facts and haveyetlostall interest exceptto
year 1924), he was always very cautious not to draw anydefinitive conclusions,
invariably historians,” for example, the epicycles of planetary motion. The mod-
qualifying his deliberations with the phrase “in the presentstate of the theory.” ern analogue of epicycles, Schrédinger prophesied, are the quantum
The fol-
lowing quoteis typical: “In the present state of the theory, it is not possible to bring
occurrence of radiative processes, not the choice between various
the jumps (1952a, 110).
possible transitions,
into direct relation with any action which finds a place in our description of phenomena, It was Born, not surprisingly, who countered Schrédingerin print.
as developed up to the present time” (quoted in Hendry 1984, 33).
‘
He defended both his collision treatment and his probability inter-
8. “The basic assumption of the individuality of the atomic processes involved at
the pretation from Schrédinger’s attack. “The most important reason,” we
sametime an essential renunciation of the detailed causal connection between physical learn, for Born’s entering into a public controversy with oneofhis “best
events” (Bohr 1937b, 18).
and oldest friends” was that Schrédinger’s publication “may have a
eseenscereras

218 Chapter Ten

Constructing the Orthodox Narrative 219
confusing effect on the mind of those who,withoutbeing physicists, are
eneinaguer:

interested in the general ideas of physics” (Born 1953b, 140). The attitude of these physicists to philosophyis truly ambivalent.
In
Born’s paper also contains a good deal of rhetoric.Its target is not a letter to Born, Schrodinger wrote that interpretive philosophical
ques-
Schrédinger’s central case against the nature of quantum jumps but tions are as important asconcrete scientific problems, only
to reverse
rather the “weakest, in fact quite indefensible, point in Schrédinger’s thjs opinion on the next page: actual scientific research is much more
arguments againstthe current interpretation of quantum mechanics”— valuable than “philosophical games’ (Schrédinger to Born, 16 May
his antiatomistic attitude. While accusing his opponentsof “having lost 1927, AHQP). Similarly, Born, after admitting that philosophicald
iffer-
the feeling of historical continuity,” Schrédinger himself, argued Born, . ences between Schrédinger and him did not matter in actual scientific
had committed a crime “that is even worse.” Schrédinger’s attemptto practice, declared that philosophical questions about the real mean-
overthrow the atomistic idea, which has proved “fertile and powerful” ing of our wordsarejust as important as the mathematical formal
8G ism .
Bo «

throughoutthe history of physics, is “almost presumptuous, and in any (1953b, 141).

case an obvious violation of historical continuity” (Born 1953b, 142). Whendo decisions on philosophical issues become important? From
Despite the sparring, Schrédinger and Born were each fully aware the exchange between Schrédinger and Bornit followsthat philosophy
that there could be no conclusive disproofof the other's position. Thus, matters when scientists want to explain and defend their theories in
front of wider audiences. Scientists then worry whether their theoreti
Schrédinger conceded, the language of quantum jumps is an extremely -
_convenient shorthandfor describing a wide range of phenomenain the cal constzucts can be interpreted in such a waythat these constructs
will .
Ee

gain acceptance in widercultural circles. Scientists aim at ensuring

atomic domain, chemistry, and thermodynamics. One can freely the
2 ees Ee Ray

talk
therefore “‘as if” these jumps occur, as long as one doesnot confusethis longevity of their conceptions,so that their work “still will be of interest
. abbreviated language with reality (Schrodinger 1952a, 119). The lan- _ other than to historians 2000 years hence” (Schrédinger 1952a, 234). _
guage of quantum jumps has acquired the status of an observational Thoughfully aware of the hypothetical nature oftheir theoretical con-
language of physics, but this does not mean that such imagery should structs, our scientific heroes (even those with positivist sympathies),
whenfaced with the harshestthreat of all— oblivion— fightfiercelyfor
re

be taken literally. The situation is similar to that of the conceptof valency,

the unobservable reality they have created. |
which is a most useful, even indispensable, shorthand in chemical re-
sk
rvs:

;
search.Initially, valency had behaved “asif” it represented a “very real Thethreatof oblivion can also come from within the scientific com-
fact of observation.” After a quantum mechanical explanation of the munity, and it was as a response to such a threat, I have argued,that
valency bond was given, valency continued to beretained, but only as the reality of quantum jumps was constituted. The concept of quantum
Se ey appara ceaeeneees

a convenient shorthand. Similarly, in the quantum world, everything jumpsnot only seemsto be a direct expression of experimental facts
happens“asif” (but only asif) jumpsare real (1952a, 119-20). (the Franck-Hertz experiments) but is a building block of the elaborate
Schrodinger himself preferred to employ different theoretical no- theoretical and interpretive framework of quantum mechanics. Yet its
tions—those that substitute a conceivable, continuous mechanism for theoretical entrenchment is no more a proofof the reality of quantum
the inexplicable andirritating “quantum jerks.” Where Schrédinger en- jumpsthantheintricate theories of electromagnetism in the nineteenth,
centuryare proofof the reality of ether.

tebe
visaged the gradual evolution of changing wave configurations, Born
witnessed abrupt quantum changes. Was Schrédinger’s imagery absurd Some kind of realism seems to be an indisputable fact of scientifi
c
life—scientists need to construct models of reality for heuristic pur-
dude A
or obviously wrong? By no means—both Born and Heisenberg, despite
their numerous objections, seemed to hold that Schrédinger’s was a poses, for a means of communication, and for describing scientifi
c ex-
feasible stand, and no conclusive arguments could be brought periments. Without some kind of realism “no communicatio
against n about
it. In fact, in an actual scientific situation “any one of facts is conceivable, even between the most sublime minds” (Born
us theoretical 1953c, 153). However, we should not project from the psychosocial
physicists, including Schrédinger, ... would use the same,or to
at least the epistemic level: though scientists must presuppose
equivalent mathematicalmethods, and if we should obtain concrete reality, they
re- need notbelievein it. Often it is such conditional acceptance without
sults our prediction and our prescription for the experimentalverifica-
tion would be practically the same.” The whole controversyis commitmentthat creates the freedom necessaryforscientific creativit
therefore y.
“not so muchan internal matter of physics, as oneofits relation to In the last analysis, Born admitted that he and Schrédinger differed
phi- merely in their conception of “the philosophical attitude of a period,
losophy and human knowledgein general” (Born 1953b, 140).
which determines the cultural background.” Born consequently
 220 Chapter Ten
Constructing the Orthodox Narrative 221
concluded that “the auspices of an agreement are
therefore frail” opposition to Schrédinger, and as springing from his desire to secure
(1953b, 150). Born’sfinal argument against Schrodinger a
wasneithersci- particle ontology, does not withstand an examination ofthe historical
entific nor philosophical, but rather social. It is an argume
nt, Born record (chapter 2). Interestingly, Born himself was awareat times of
noted, that Schrédinger himself “is not too proud to
use” (1953b, 149)— his lapses of memory.’ The lapses, of course, are not cases of memory
quotation of authorities. Born chose to quote from lette
r from Pauli, simply failing but of memory selecting. It is natural in retrospect to see
“whois generally acknowledged to be the mostcritical,
logically and later developments in earlier ideas, to forget the mistakes, the ambiva-
mathematically exacting amongst the scholars who have
contributed lence, to suppress the false starts, and to bring to the fore the valid,
to quantum mechanics.” In this letter Pauli described
Schrédinger’s accepted aspects ofthe story. Scientists turned historians are natural
effort as “a dream of a way back to the classical
style that seems fo » “rational reconstructionists.” And even the authors of discoveries (and°
me hopeless, off the way, bad taste... and not
evena lovely dream” perhapstheyespecially) are “Whiggish”historians. In this way, tenta-
(1953b, 150).
Schrédinger must have loathed such arguments, especi tive early attempts are described as if they had alreadyentailed the in-
ally those tellectual preferences that only formedlater.
that cited the consensus of physicists against his interpretati
on. “Since Born’s probabilistic interpretation in terms of particles applied ini-
whenarescientific questions decided by majority?” he would ask Born
tialiy, if at all, only to free particles. And for free particles (except for
(quoted in Born 1961, 87). In his obituary for Schrédinger, Born
‘gave the, special case of the harmonic oscillator—Schrédinger 1926e) wave
his answer—“Atleast since the time of Newton” washis final reply
packets disperse, as Heisenberg, Born, Lorentz, and even Schrédinger
(1961, 87). himself understood,” and the Born-Heisenberg particle interpretation
is the mostnatural. Yet, as I discussed in chapter 2, the crux of Born’s
Indeterminism andHistoriographical Doubts -
efforts lay elsewhere: in describing quantum mechanically the states
Today physicists and philosophersof quantum physics conside and energetic changes of atoms. And for this—the more interesting
r entan- and more complicated case of the bound system—the matter wavein-
glement,rather than acausality, the mostdistinctive featureof
the quan- terpretation a la Schrédinger is the more natural, as Bohr emphasized
tum world. Indeterminism turns out.to be less interes
ting than non- from the beginning (chapter 6), and as both Heisenberg (1958) and Born
locality: one can have both deterministic and stochastic hidden variabl
e (1969) later conceded. The reason Born suppressedthis factin his recol-
theories, yetall of them must be nonlocal (Homeand Whitaker 1992).
In lections, singling out only the more obviouscaseof colliding electrons,
the previous chapter I analyzed the rhetorical strategies used
by the is that Born’s history is an ideological one, emphasizing particlehood
foundersof the Copenhagen interpretation to endow determinism
with in contrast to Schrédinger’s wave ontology —a threatening alternative
“finality.” In this chapter I have argued that Bohr’s histori
cal recollec- to the Géttingen-Copenhagen version of the quantum world. This ide-
tions distort history to the same end. Born,the originatorof
the proba- ology was reinforced by the deception that the particle interpretation
bilistic interpretation, similarly constructed the course
of his recollec- wasthereall along, and inevitably so. Together with indeterminism,it
tionsto lead naturally to the “inevitability” of indete
rminism. formed “a natural interpretation” in Feyerabend’s (1975) sense. This is
That one cannot trust the recollections of participants in events
is a a commontechniqueofusing history for propaganda. The Copenhagen
historiographical platitude. As Einstein himself put it: “Every
reminis- spirit was diffused by accommodating history to ideology. Because the
cence is colored by today’s being whatit is, and therefore by
a deceptive opponents (Schrédinger, Einstein) did not write detailed recollections
point of view”(1949a, 3). There are indeed many factual
discrepancies on this topic and because manyhistorical studies take the map drawn
and confusionsof dates in Born’s recollections of past
events. Thus Born
described his work with Norbert Wiener (Born and
Wiener 1926) as
having been doneafter his interpretation of the wave 9. “Doch ist meine Erinnerung und die Ereignisse vor 35 Jahren etwas verblasst”
function (inter- (However, my memory of events {that took place] thirty-five years ago is somewhat
view with Born, AHQP). There are also obvious distort
ions: in his No- faded) (Born 1961, 85).
bel lecture Born (1955b) recalled his opposition to Schrédi
nger’s ideaof 10. Heisenberg argued this point against Schrédinger’s wave ontologyin his uncer-
a weight function, while in his original papers Born (1927b) tainty paper (Heisenberg 1927b; see chapter 4) and Born did so ina paper published in
perceived
this idea of Schrédinger’s as being close to his own. Nature (Born 1927b). For Lorentz’s and Schrédinger’s discussion of the dispersion of wave
his y . - Born’s accountof
is interpretation as having been conceived from the very beginning in
packets, see Lorentz to Schrédinger, 27 May 1926, and Schrédinger to Lorentz, 6 June
1926, both in Przibram (1967).
 222 Chapter Ten
ame kt aN Ni aehn na rata esaan ate

by the Géttingen-Copenhagen physicists as a faithful guide to pivotal

developments, one-sided and biased accounts are formed CHAPTER 11
.
Ideological history has its price, of course. By concentrating on |
the
more straightforward interpretation of IC, |? asa density
of scattered
particles, Born omitted the most problematic, yet most valuabl
The Myth of Wave-Particle Complementarity
e partof,
his contribution (the probabilistic interpretation ofstationary states):
the
core around which pivotal developments would take place (chapter
2). During the summerof '25 .. . | asked him [Pauli] this question: “Waslight[a]
Born’s probabilistic interpretation was, as I have argued,
a process wave or a particle?” And I can rememberthis kind of cynical laugh ofhis.
in flux, during which Born’sintellectual preferences formed
gradually,
while his enthusiasm for Schrédinger’s theory faded (chapter Interview with Hendrik Casimir, 28 January 1964, AHQP
2). There
was heavypressure from Born’s younger colleagues (especi
ally Heisen-
berg) not to succumb tothe seductive classical beauty of
Schrédinger’s It ain't exactly real, or it is real, but ain't exactly there.
theory. This pressure was notthe only and perhaps not even the
main
reason for Born’s about-face on the physical significance of Schréd Leonard Cohen, “The Future”
ing-
er’s approach. The realization that wave packets disperse, the
insight
that position space does not have a real ontological preference, and This idea [of wave-particle synthesis] seems to me so natural and simple, ~
the
brilliant extension of Born’s ideas by Pauli, Jordan, and to resolve the wave-particle dilemmain such a clear and ordinary way,
Dirac—all are
substantial reasons for Born’s changeof mind.Still, the direction of thatit is a great mystery to me that it was so generally ignored.
the
modifications in Born’s ideas and Born’s later description
of the devel-
opmentof his ideas are not fully comprehensible without taking
J. S. Bell 1987, 191
into
accountthe Géttingen-Copenhagenalliance against Schrédinger.
Born’s
militant public stand on the issue of indeterminism after
1927 (despite
his private concession to Einstein that this issue is undecidable)
also
belongsto the ideological partofthis story.
Introduction: The Dramatic Historical Narrative-
Before 1927, neither Born northe other G6ttingen-Copenhagenphy
s-
icists took anydefinite stand on the issue of indeterminism. Althou
gh The story of the development of quantum mechanics builds dramati-
they wereall opento the possibility of an ultimately probabilistic
phys- cally around the wave-particle dilemma. We follow the piling up of
ics, this openness did not imply preference. The transition to a
“com- contradictions between the corpuscular and wave aspects of matter and
mitted” indeterminism was conditioned, for Born
and Heisenberg, radiation, one after another, until the resolution of the ultimately un-
by theoretical possibilities and competitive dialogues overcon
flicting bearable conceptual tension by Bohr’s principle of “wave-particle com-
tastes in interpretation withinthe scientific community. “Comm
itment” plementarity.” This exciting narrative appears in many textbooks (Born
is perhaps too strong a characterization. Thus at times
Heisenberg 1969), philosophical accounts (Sklar 1992), and personal recollections
(1931la, 1931b) adopted a relaxed attitude toward the
causality issue, (Heisenberg 1958). Thomas Kuhn singled out wave-particle.comple-
claiming that by an appropriate redefinition of causali
ty, its validity mentarity as a prime example of a revolutionary paradigmatic transfor-
couldstill be maintained.
mation (Kuhn 1970, 11-12).
The photoelectric effect and the Compton effect, so we often read,
constitute direct proof of the existence of light quanta: these experi-
ments can be accountedfor only by assuming the corpuscularnature of
radiation.! Similarly, the Ramsauer and Davisson-Germer experiments,
and the theoretical contributions of de Broglie and Schrédinger, point

1. “The corpuscular natureoflightitself is proved in the most obvious way bythe laws
of frequency changein the scattering of X-rays [the Comptoneffect]” (Born 1964, 87).
 224 Chapter Eleven
The Myth of Wave-Particle Complementarity 225
unequivocally to the wave nature of matter. These develop
ments, sup-
posedly, caused acute distress, for The dramatic narrative assumes, and naturally leads to, the thesis
they “seemed to eontiadict eat
other without any possibility of compromise” that wave andparticle concepts are mutually exclusive, and to a reso-
(Heisenberg 1958 37)
The unhappyquestion was: “How could the same thing lution of the “paradox” through Bohrian concepts of complementarity.
Pee Was igo eee 1958, 35). “How could this be a wave and Yet the majority of working physicists did not consider wave andpar-
be understood?”
ar (Bom
Sled?” ) 186,54)
. “How
, are those c ontradii ctory aspects to ticle characteristics irreconcilable. Even William Henry Bragg, who in
'
be recon-
anecdotal accounts of history canonized the incompatibility between
Whenweread papers dealing with the wave-pa waves and particles by saying that physicists use particles on Mon-
rticle issue before the
rise of the Copenhagen philosophy, we hardly days, Wednesdays, and Fridays and waves on Tuesdays, Thursdays,
find feelings of despera-
tion or distress. A patient suspense of and Saturdays, was not“averse to a reconcilement of a corpuscular and
final judgment seems to be a
more fitting characterization of the attitude a wavetheory.” Instead, Bragg said: “I think that somedayit [recon-
of physicists? Toprovide
a unified theory of matter and radiation seemed cilement] may come” (quoted in Wheaton 1983, 166 n). The work of
creative physicists—Einstein, de Broglie
more ieaportent to Einstein and de Broglie was inspired by the goal of finding similar
, Jordan, Born, Debye, Dirac
Pauli—than to establish the ontological status of foundations for the structure of matter and of radiation; Born’s work
their constructs. THE was likewise guided bythe search for a unified description of matter
wise resignation from the pursuit of ultimate answer
s allowed these and radiation. Born conceived his probabilistic interpretation of the y-
physicists to borrow freely from each other's.
powerful techniques, function in analogy with the connection Einstein had proposed be-
even when their conceptual presuppositions
seemed fundamentall ,
opposed to each other(Darrigol 1986, 198). tween the intensity of the electromagnetic field and the density oflight
* quanta. Born (1926c)said explicitly that he hoped such an approach,
Impressive theoretical advances were possibl
e, despite the wave-
particle dilemma, becausethe interpretation if successful, would be a step toward a unified theory of matter and
of “crucial experiments”
was not constrained in one particular way, radiation.
as the received narrative
implies. There were many non-Einsteinian explanations While it remained unsolved, the wave-particle “paradox” was pa-
of the photo- tiently put aside in the hopes that eventually, when a consistent atomic
electric effect that ingeniously avoided the hypothe
sis oflight quanta
(Stuewer 1970).3 Similarly, there were alternative explana mechanics was found, the problem would resolve itself. For Heisen-
Compton effect—a prominent alternative to light
tions of the berg, the powerful new transformation theory accomplished this task
quanta was provide
d and allowed a particle interpretation in the spirit of a generalization of
by the Bohr-Kramers-Slater theory, as I have
discussed earlier (chap-
ter 2).It was in order to avoid Einstein’s concept oflig Born’sstatistical interpretation. In Heisenberg’s opinion, Bohr, who did
Bohr introduced “further peculiarities” into the space-tht quanta that not participate in the mathematical development of the new quantum
ime description theory and did not experience firsthand its wonderful power, remained
of the microdomain, while dealing with the Compto
neffect in terms of trapped in the visualizable language of wave-particle duality.4 As Hei-
a virtual radiation field. Bohr did not accept even
the Bothe-Geiger ex- senberg putit: “These paradoxes were so in the center of his [Bohr’s]
periments as crucial evidence for light quanta (chapter
6), thoughlater mind that he just could not imagine that anybody could find an answer
constructing a narrative of wave-particle complem
entarity, he would
cite the Comptoneffect as “proving” their existen to the paradoxes, even having the nicest mathematical scheme in the
ce. Nor did the wave
aspect of radiation unequivocally constrain theorists world”(interview with Heisenberg, 25 February 1963, AHQP).
to wavelike ex- While rejecting Schrédinger’s vision of a pure waveinterpretation,
planations. Papers by Duane, Epstein and
Ehrenfest, and Jordan inter-
preted the diffraction of material particles by Heisenberg was equally reluctant to acknowledge wave-particle du-
crystals purely in terms
of particles (Darrigol 1992b). (In fact, these ality as unavoidable. Thefact that one needed contradictory, wave and
papers could be interpreted
as providing a physical explanation of how particle, concepts for the intuitive interpretation of quantum mechan-
crystal periodicity might
producetheillusion of the wavenature of the electron ics, Heisenberg thought, merely indicated that something essential was
.) 6 still missing from the present formulation of quantum theory (Heisen-
2. See Wheaton (1983) for a description of the
3. spanHs
early treatment of wave-particle duality. berg 1926b, 993; these words were written before the developmentof
with a typical statement: there “is absolutely
countin
no possib le way of ace
THe 10 g for P photoelectric effect.
ec .. except by adopting
i i
the idea of the photon - (Tolansky-
-
4. “Bohr had not hadthis experience, so I would say he was not so much impressed by
the new developments”(interview with Heisenberg, 25 February 1963, AHQP).
 The Myth of Wave-Particle Complementarity 227
226 Chapter Eleven
one has to use “two in-
tion, for the mathematically uninitiated, that
transformation theory). Heisenberg continued to be critical of wave- complete analogies—the wave pictu re and the corpuscular picture”
particle duality even after he, yielding to Bohr’s pressure, appendedto (Heisenberg 1930, 11). These analogies
are “accurate only in limiting
the uncertainty paper the postscriptin Bohr’s spirit. “If one begins with cases.” We use them in a limit ed way, becau se we attempt “to describe
a statement ‘there are waves and particles, everything can be made things for which our language has no words
cle duality is
.” Wave-parti
phy
contradictory free” was his disapproving sketch of Bohr’s philoso of our own making:“light and matte r are singl e entities and the appar-
in a letter to Pauli written in the spring of 1927 (chapter 6). our language” (1930, 10). An ade-
ent duality arises in the limitations of
As I have argued, contrary to the usual historical account that the particle concepts “cannot
in quate analysis of the limitations of wave and
heated controversy between Bohr and Heisenberg eventually ended the mathematical apparatus
be carried through entirely without using
complete agreement due to Pauli’s skillful intervention, a genuine una- of quantum theory” (1930, 11-12). Nor
can the mutual exclusiveness _
berg ly maintained in thought
nimity of opinion between the two men never occurred. Heisen
of wave and particle concepts be consistent
stated clearly in an interview with Kuhnthat he never accepted Bohr's experiments. In an analysis of the y-ray
thought experiment, lightis
dualistic approach.This is the reason the papers of Jordan and Wigner treated as a wave anda particle simul
taneously. As Heisenberg putit:
(1928) on second quantization made him very happy—they demon- “Tt is characteristic of the foregoing discu
ssion that simultaneous use
do
strated not that one needsboth waves and particles but that one can is made of deductions from the corp uscu lar and wave theoriesoflight,
erg emphas ized, “one mathema ti- r, and, -on the other
it “either way.” “We have,” Heisenb for on the one hand, we speak
of resolving powe
one mathe-
cal scheme that allows many transformations, but . . . just (1930, 22-23).
... is hand, of photons and their recoils”
- matical scheme... . The fact that we can use two kinds of words. The above sent ence revea ls why Heisenberg was so reluctant to
And whenK uhnas ked paper. It
just an indicat ion ofthe inadequ acyof words.”
y to Hei- adopt Bohr’s “correct ion” to his discussion in the uncertainty
about the y-ray thought experiment, in which Bohr—c ontrar in his analysis of simple thought
wave fea- also explains why Bohr moved swiftly,
senberg’s first, mistaken account—used both particl e and hragms, abandoning the ca-
explain ing the y-ray ex- experiments, to rigid and movable diap
tures of radiation, Heisen berg insisted : “For this experimentviolates consistency
nonical y-ray thought experiment:
periment, it was useful to play between both pictures. . . . But
it was not of contradictions in the use of
indepen- as defined by Bohr, who held that Jack
absolutely essential. You could actually use both langua ges ensu red by mutual exclusivity of
wave and particle concepts is only
dently” (interview with Heisenberg, 25 February 1963, AHQP). experimental arrangements. Wecan also
understand why mathematical
Heisenberg’s recollections are confirmed by his work at the time, es- physicists, such as Bopp and Heis enbe rg himself, maintained that one
pecially his Chicago lectures (Heisenberg 1930). While these lectures cannot obtain a contradiction-free
interpretation of quantum theory
were intended to spread the Copenhagen spirit and are often interpr
eted on the concepts of classical phys-
rticle com- through thought experiments that rely
as endorsingall of the Bohrian doctrines, including wave-pa
al dis- ics, such as wavesandparticles.
plementarity, a careful reading finds that Heisenberg’s princip
ced
agreement with Bohr persists. What Heisenberg proudly announ Mathematical Physicists and the
Wave-Particle Dilemma
was not wave-particle complementarity but rather, contra Bohr, sym-
metry, a “complete equivalence of the corpuscular and wave concept
s,
ndational myth” of quantum
Wave-particle complementarity is a “fou
which is clearly reflected in the newer formulations of the mathematical By calli ng it a myth, I do not deny
mechanics (Dineretal. 1984, vii).>
theory.” While taking a stand at odds with Bohr’s, Heisenbergpaid lip ibly philosophical), problem
that there is a genuine physical (and poss
service to Bohr’s philosophy: there is nothing in the book that is not of matter and radiation. Yet
connected with the wave-particle duality
contained “in previous publications, particularly in the investigations is, and its alleged definitive
thevery construction of what the problem
of Bohr”; the purpose of the book will be fulfilled if it contributes to the ous, ignoring much of the
resolution by complementarity, is tendenti
diffusion of Copenhagener Geist der Quantentheorie (1930, preface). Such this issue .
fascinating, research connected with
statements contributed to the mistaken conflation of Bohr’s and Heisen-
berg’s positions, andto a furtherproliferation of contradictions in Hei- Ein-
emphasized,in the spirit of de Broglie and
senberg’s presentations. 5. The editors of these research papers the “mutual exclusivity,” of
1984, vii), rather than
The mathematical scheme of quantum mechanics “seems entirely stein, the “coexistence” (Diner et al.
these concepts.
adequate for the treatment of atomic processes.”It is only for visualiza-
 228 Chapter Eleven
The Myth of Wave-Particle Complementarity 229
Philosophically, it can be argued that the ortho
dox solution is deeply particles in quantum mechanics is characterized by a “peculiarity”
unsatisfacto ry. To say that microscopic objects exhibit wave
Properties because they obey a “complemen andparticle similar to a wave theory, as Schrédinger, Pauli, and Jordan early
tarity between waves and real-
particles” (Bohr) or because they are “wavicles” ized—the so-called interference of probabilities (Jordan 1927a, 647).
(Bohr’s carele ss fol- Thus,in the two-slit experiments, the probability of a particle’s passing
lowers) is not muchbetter than to say that
sugar is sweet becauseit has thrdugh the first slit when the second slit is closed is |y,|2, while the
an essential qualityof “sweetness.” The foundationalstatus of wave-
particle complementarity rests on Bohr’s probability of its passing through the second slit whenthefirstslit is
controversial, and as I have closedis |y.|?. The probabilistic pattern when bothslits are openis not
argued unfounded, doctrine of the necessity of classical
gives up the foundational status of the later
concepts.I f one iw,|? + |[w]?, as it would be in the case of classical particles, but |y, +
doctrine, the foundational Wa |2—an expression that, when expanded, contains a supplementary
status of wave-particle complementarity disso
lves as well. The wave- “interference” term in addition to ||? and | y,|?.
particle dilemma consequently becomes a prob
lem for experimental in- This.situation is completely analogous to an explanation of Thomas
vestigation—the way,infact, it is treated in the recen
t physicsliterature Young's two-slit experiment whenthe difference of phases adds up in
(Ghose, Home, and Agarwal 1991, 1992:
Ghose and Home 1992; Scully , a “constructive” and “destructive” manner to produce a typical inter-
Englert, and Walther 1991).
If one gives upBohr’s prohibition on deducing featu ference pattern.” Asit is in optics, it is not the probabilities that sum
res of quantum up, but rather the amplitudes—in a general formulation the particle
reality from the characteristics of the mathematical
formalism, a rich and wave properties apply to every quantum mechanical process. One
varie ty of solutions to the wave-particle dilemma
become possible, can, if one likes, maintain that both wave and particle attributes are
depending onthelocal theoretical problem oneis faced
with. Mathe- in some sense essential to quantum mechanical description. Onecan,
matical physicists, from the very origins of quantum
physics, suggested. however, consider the particle substratum morebasic, maintaining that
different solutions to the wave-particle dilemma,
connected with the the interference of probabilities is merely a formal analogue of a wave
characteristics of their mathematical theorizing,
though often being description, expressing the nonclassical nature of the essentially cor-
carefu l not to take definitive interpretive stands. Mathe
maticalresults puscular quantum world (Landé’s position—see the discussionlater in
teased the imagination of physicists in more wayst
han one.
Dirac, whose outstanding contributions shaped mathematic No wonder this chapter).
tions of the wave-particle issue, felt, contra Bohr,
alelabora- Proponents of the Copenhageninterpretation vacillate in their writ-
that “the final word . ings between these two options. Such vacillation contributes to con-
wasnotyet said about wavesand particles” (1963,
53). ceptual confusion aboutthe wave-particle issue. The particle-kinematic
In the context of nonrelativistic quantum mechanics, physic
ists often probabilistic interpretation, which is based on the presupposition that
relate to Born’s corpuscularstatistical interpretation,
and its extension it is meaningful to talk about the probability of a perfectly sharp posi-
by transformation theory, as a solution of the wave-
particle dilemma: tion at a specific time, is not compatible with theliteral applicability of
a subst ratum of particles is guided by wavelike probab
ilisti
c lawful- de Broglie—Schrédinger matter wave packets in three dimensions,as is
ness.’ Yet this interpretation is incompatible with Bohr’
tual exclusivity. As
s thesis of mu- _- often discussedin the orthodox writings. As Born himself pointed out,
Born putit: “Both pictures are necessary
for every
real quantum phenomenon”(1964, 105).* Pauli and a particle represented by a wave packet “cannot have any sharply ba
Born, whenthey did fined position, since it is only a group of waves with vague limits
not publicly support Bohr, interpreted wave-parti
cle duality through " (1927b, 9). A more consistent stand would be to deny the “reality” of
Born’s probab
ilistic interpretation, with the associated
mathematical matter waves in the three-dimensional case as well (the Gottingen-
notion ofthe interference of probabilities. Such
a statistical interpreta- Copenhagen physicists denied the reality of Schridinger’s waves in
tion “prevents contradictions between wave mecha
nics and the me- the 3N-dimensionalcase), confining all discussions to a probabilistic
chanics of point particles” (Pauli 1973, 13).
The statistical calculus of
6. De Broglie: “The present theory should be taken 9. The expression |¢,(qo)|?dq is the probability that a coordinate lies in the interval
as a form whose physical content (95, 9o + 4q) when theatomis in the nth stationarystate. In general, let us denote by |P(Bo.
is not entirely given” (quoted in Darrig
ol 1986, 209). One finds similar expres
Papers byBorn,Einstein, and other physici sions in qo)|?dq the probability that q lies in (qo, 49 + dq) when another mechanical variable has
sts.
7. Maddox (1988) calls it a “delicious compr the value @. Similarly, the probability that Q takes a value in (Qo, Qa + dQ) while q has
omise” between Bohr and de Broglie,
8. Born’s public supportof Bohr’s position result the value qp is given by | (qo, Qo)|?dq. Then the probability that Q lies in the interval (Qo,
ed in the confusionand inconsisten- Qn + dQ) while B = Bois given not, asin the classical case. by dQ-f!(Bs,q)|?-|u(gq,Qo)|2dq
cies that characterize Borm’s writings on the wave-p
article issue. but rather by dQ-|@(8o, Qo)|?, where ®(Bo, Qo) = [6(Bo, qW(G. Qa )dq (Jordan 1927a, 647).
 230 Chapter Eleven
The Myth of Wave-Particle Complementarity 231

treatment in terms of particles. Yet the appeal of wave imagery is too In 1925-27, that period of “human confusion,” support for Schré-
powerful andits utility for “pedagogical” purposes is too appealin
g, so dinger’s position came from unexpected quarters. Though he attacked
even Heisenberg and Pauli often preferred to use the matter wave pre-
Schrédinger publicly for deviating from, and challenging, Bohrian
sentation of individual particles. (See the discussion later in
this guidelines (especially on the issue of quantum jumps), Jordan’s theo-
chapter.) . retiéal efforts were muchcloser in spirit to Schrédinger’s than to Bohr’s
As opposed to the partial analogy between waves and particles con- ideas.° As noted by several authors, matrix theorists were familiar
tained in the idea of a wave packet, the Born-Pauli-Jordan idea
of the with, and even actively participated in, an elaboration of the wave theo-
interference of probabilities allowed an overarching analogy between
retical point of view (Beller 1983; Darrigol 1986; Kojevnikov 1987; see
matter and radiation. The fact that the interference of probabilities
in also chapter 2). Jordan, a founder of second quantization, is the most
Schrédinger’s formulation correspondsexactly to the rules of multipli
- prominent case. Following de Broglie and Einstein, Jordan made many
cation for matrices withinfinitely many rows and columns confirm
ed important advances using the assumption of a complete analogybe-
the great generalityof thisnew probabilistic framework of quantum
me- tween matter and radiation as a heuristic principle. In Jordan‘s hands,
chanics (Jordan 1927a, 647). All existing versions of quantum mechan-
the fruits of the matter wave analogy included a theoretical proof of
ics—Heisenberg’s matrix mechanics, Schrédinger’s wave mechani
cs, electron diffraction—before the experiments by Davisson and Germer
Born-Wiener’s operatorcalculus, and Dirac’s e-number calculus—were
that demonstrated the diffraction of an electron beam—as well as
united in the single, powerful formulation of probabilistic transfor
ma- Jordan’s early conjecture that matter could be created and destroyed
tion theory: That the theory was probabilistic did not seem too high a ' (Darrigol 1986, 219). ;
price comparedwithits assets of unity, generality, and problem-solving As early as the Born-Heisenberg-Jordan paper, Jordan had obtained
power. The unity and generality of a comprehensive probabilistic quan-
light quanta by quantizing the Maxwell fields (this part of the paper
tum theory was the most powerful argument for the. indeterministic
was written by Jordan alone). As soon as Jordan learned about Schré-
character of the quantum world and for the duality expressed in the dinger waves, he contemplated the possibility that matter as well could
interference of probabilities. For Heisenberg, all his previous intuiti
ve be represented by quantized y-functions. His presumption of an anal-
discussions with Bohr aboutthe duality of waves and particles were at
ogy between matter andradiation led naturally to the idea thatclassical
once superseded bythe unity and beauty of the newborn mathematics.
individual particles might be replaced by wave excitations in the case
Schrédinger thought otherwise. Since 1925, whenhe became familiar .
of matter as well. Jordan had very little encouragement from Heisen-
with de Broglie’s and Satyendra Nath Bose’s ideas, he hadrealized that
berg; Born, initially sympathetic to the idea, soon withdrew his support
Bose-Einstein counting undermines the idea of an individual distin- (Darrigol 1986, 219). .
guishable particle. Particles ina gas hadto be replaced by “energy
ex- Jordan’s quantized matter waves in three-dimensional space rein-
citation states” (Schrédinger 1926g), and the corpuscular picture of a
forced Schrédinger’s wave theoretical position, by removing the multi-
gas hadto be replaced by de Broglie waves. This insight spurred Schré-
dimensionality of the y-function as an argument against Schrédinger’s
dinger’s theoretical efforts, culminatingin the crowning achievementof
wave ontology. There was a meaningful similarity between Jordans
the creation of wave mechanics (Wessels 1979). For the rest of his
life matter waves and the quantized waves Schrodinger used to explain
Schrédinger, considering the concept of an individ
ual particle bank- Bose’s statistics before his discovery of wave mechanics.If for Heisen-
rupt, advocated an underlying wave ontology, rather than the synthe-
berg second quantization implied the equivalence, or mutualtranslat-
sis, or “coexistence,” of waves and particles in the spirit of de
Broglie ability, of the wave andparticle models, for Jordan the wavetheoretical
and Einstein (Bitbol 1996; Beller 1997a). Understandably
, Schrédinger substratum was the morebasic of the two. Like Schrédinger, Jordan
did not embrace the corpuscularstatistical ontology offered
by Heisen- was willing to give up the primacy of atomistic ideas. In this theoreti-
berg and Born.
Heisenberg also realized very early that the Bose-Ei cal context Jordan heldthat the discontinuity of both matter and light
nstein statistics can be derived from wave quantization: “The existence of atomsis
underminesthe idea ofan individualparticle and used
this conclusion no longer a primary basic fact of nature; it is only a special part of a
initially to argue against the possibility of intuitiveness,
or Anschaulich-
keit, in the microdomain (Heisenberg 1926b). He later 10. Jordan’s theoretical work, and his crucial role in the history of quantized matter
minimized the im-
plications of the Bose-Einstein statistics in his interpre waves, is analyzed extensivelyin Darrigol (1986).
tive writings when
confronted with Schrédinger’s competing, wave theoretical 11. Born may have withdrawn his supportbecauseofhis change of heart regarding the
endeavors. physical significance of Schrédinger’s wave ontology(chapter2).
 232 Chapter Eleven
The Myth of Wave-Particle Complementarity 233
much more general and comprehensive phenomen
on— the phenom- the joint applicability of position and velocity variables as
enon of quantum discontinuities (1944, 144-4 belonging
6, quoted in Darrigol to the very definition of a particle. For Schrédinger, the indistinguish-
1986). Schrédinger, as we know, did not perceive secon
d quantization ability of particles, implied by the new quantum statistics, signifie
as implying equivalence between wave and partic d
le pictures in the the total bankruptcyofthe conceptof a particle, and the continuing use
spirit of Heisenberg, for by this procedure one “cann
ot avoid leaving of particle concepts offendedhis theoretical sensibilities. What Schro-
indeterminate the number of particles dealt with.
It is thus obviousthat dinger found so objectionable in the idea of a particle was that “it con-
they are not individuals” (1950, 112).
The fascinating history of wave-particle concepts canno stantly drives our mindto ask for information which has Obviously no
t be fullyun- significance... . - An adequate picture must not trouble
derstood without introducing the history of quantum us with this dis-
electrodynamics quieting urge; it must be incapable of picturing more than there
(QED), whichis beyond the scopeofthis book. Let me simpl is.”
y mention According to Schrédinger, the problem with the conceptof a particle
two importantpoints. First, different scientists deduced
different philo- wasthatit “exhibits features which are alien to the real
sophical conclusions from the same, or very simila Practice” (1950,
r, mathematical pro- 111). In the microdomain,the only “tolerable image” of even an isolated
cedur
es. While not discussing these issues explicitly, their
experience
and knowledge of QED informed their popular writi particle is “the guiding wave group” (1950, 115). For Bohr, this was an
ngs, which were indicationof the limitation of classical concepts, not an indicationthat
confined, as it were, to the nonrelativistic theory
of quantum mechan- they should be given up. Feynman’s verdict wascrisper: “It [the elec-
ics. This state of affairs produced contradictions,
most prominentlyin tron] is like neither” (Feynman, Leighton, and Sands 1969, 37-1).
Heisenberg’s case. The second point is that even for
the same author, Difficulties in the discussion of the wave-particle issue were further
the relative primacy given to particle versus wave conce
pts changed aggravated bydisagreement about which (wave or particle) attributes
from one theoretical context to another. Thus, in their
early compre- are “essential” and which are merely artifacts of interaction.12 Here am-
hensive program for QED, Pauli and Heisenberg (1929) assum
ed the biguity was connected with a fruitful theoretical freedom, exploited by
“symmetry,” or equivalence, of wave and particle schem
es. Later, dis- physicists from the early days of quantum theory until today. We have
appointed by the difficulties they encountered, Pauli
preferred to em- already mentioned that Duane, Jordan, and other physicists did not feel
phasize the dissimilarity between matter waves andlig
ht waves(matter compelled to consider the diffraction of light (or matter) to be an indi-
wavesare not directly observable—y is a symbolic quanti
ty). Similarly, cation of its wave nature—thediffraction was merely anartifact of the
Dirac, whoseefforts were initially guided by a prefe
rence for particle quantized structure of the grating used in the experiment. Similarly,
concepts,in his later years gave much more support to the symme
trical theorists who wished to avoid ascribing discontinuities to radiation
interpretation of second quantization (Darrigol 1986).
Important advances using second quantization were possi (Debye and Ehrenfest—see Darrigol 1986) located all the discontinui-
ble with- ties in the interacting matter.
out a resolution of the wave-particle dilemma: Fermi and
Fock, for ' Even localized detection of light quanta on a photographic plate,
example, used formal theoretical arguments
in their work (a charge- though generally held to be an unequivocalsign of their Particlehood,
Symmetry argument, an argument of nonconserv
ation of particle need not be considered so—localized detection events “can be regarded
number). In conclusion, no paradigmatic conse
nsus was achieved, and as originating from the quantized energylevels of the atomic constitu-
indeed none was needed, to ensure outstandin
g theoretical progress ents of the detector” (Ghose and Home 1992, 1438).!3 What some quan-
(Darrigol 1986; Schweber 1994). No unchanging philosophi
cal meaning tum theorists consider today a more adequate indication of single
can beattached to such creative theoretical efforts,
even if these efforts particle-like behavior are “single photon states” of light.For an ideal
can be reconstructed as being directed toward the same
foundational
issue—the wave-particle dilemma.
12. From a verificationist point of view, this distinction is meaningless, yet theoreti-
Ambiguity and the Wave-Particle Issue cally it is difficult to abstain from discussing abstract cases. Thus Bohr often talked of the
propagation of radiationin free space, which,in his opinion, must bepictuted by a wave
theoretical model.
Theproliferation of opinions, or perspectives, on the
wave-particleis- 13. “One observes discrete events giving rise to an interference pattern even with
sue was connected with the ambiguity of the designatin
g terms. There strongly attenuated sourcesof classicallight” (Ghose and Home 1992, 1438),
was no agreement on the necessary and sufficient 14. Such “single photon states” are Fock-spacestates that are eigenstates of the “pho-
attributes of a par-
ticle. Einstein and Schrédinger argued thatit is unrea ton number operator,” which corresponds to the eigenvalue unity (Ghose and Home
sonable to give up 1992, 1436).
 234 Chapter Eleven
The Myth of Wave-Particle Complementarity 235

single photon state, the probability of joint detection of more than

It is not clear why two points are necessary for the definition of a
one photon is zero—single photon states imply the notion of single
particle, unless one intends with these two points to specify the par-
particle-like behavior.With this definition of particlehood, Ghose and
ticle’s path—something forbidden by the Copenhagen version of quan-
Home argue, Bohr’s thesis of mutual exclusivity is contradicted: one
tumtheory. Born, in fact, inconsistently talks about particle paths: “If
can have experiments in which single photonstates exhibit a tunneling
we propose to carry out the ‘demonstration of a corpuscle,’ we must
effect (tunneling is considered exclusively a wave phenomenon). Other
settle the question whether its path has gone through the upper or the
physicists argue the applicability of Bohr's thesisof mutual exclusiv-
lowerofthe twoslits of the receiver” (1969, 101, myitalics); or similarly:
ity by considering interference, rather than tunneling, as the hallmark
“Tf pure interference is to be observed, we are necessarily precluded
of wavelike behavior (Scully, Englert, and Walther 1991).'° Clearly, the
from making an observation of any point of the path ofthe light quantum
lively discussion continues. The thesis of the mutual exclusivity of
beforeit strikes the screen” (1969, 102, myitalics).?” -
waves and particles is found applicable to some situations, yet inade-
The frequent use of the notion of the “path”of a particle, ora particle :
quate in others, depending on the definitions of the terms used.
“trajectory” (Pauli 1973, 14), in the discussion of such thought experi-
Theoretical preferencesrelating to the wave-particle issue are closely ments— despite the Copenhagen claim that continuous space-time mo-
connected with intellectual temperament and personal research expe-
tion is incompatible with the uncertainty relations— demonstrates that
rience. Matters of taste often dictate what seems fundamental and what such visualizable interpretations cannot be carried through w ithout
derivative. De Broglie, considering quantization a mystery, introduced
inconsistencies, just as Heisenberg and Boppasserted. And with this
wave-particle duality as a way to deduce and elucidate Bohr’s quan-
realization, wave-particle complementarity, as Bohr conceivedit, loses
tized energy levels in an atom. It was a seductive project: even Bohr, much of its power and appeal. —
despite his claim of further “irreducibility” of the “atomic postulate,” Bohr had his own definitions of wave and particle attributes. He
initially praised the de Broglie~Schrédinger approach forits ability to never accepted a complete analogy, or basic symmetry, between matter
decode quantization (chapter 6). Others, such as Duane, took quanti- and radiation, maintainingall his life a peculiarly unshakable commit-
zation as given and constructedtheir views about the wave-particle is- ment to classical wave theory and somewhatresenting assertions of the
sue accordingly. / reality of light quanta. Thus Bohr claimed that “there can beno ques-
The ambiguity of wave and particle terminology resulted not only in tion of replacing the wave picture of light” (1933, 5) and that interfer-
theoretical freedom butin an abundance of ad hoc moves and inconsis-
ence patterns offer so thorougha test of the wavepicture . . . that this
tencies. Born’s case is typical. Discussing the Bohrian doctrine of com- picture cannot be considered as hypothesis” (1933, 4). When one deals
plementarity, Born defended the thesis of the mutual exclusivity of with the interference picture, “light quanta with such definite entities
waves and particles by an ad hoc extension ofthe definition of a particle
of energy is something which of course does not come into the picture
(a discussion strangely at odds with Born’s own probabilistic interpre- at all. It is just that whereverlight energy is released into material, then
tation of quantum mechanics). Claimingthatit is misleading to assert it isin the quantity hv” (1937d[lecture 3], 269; see also Stachel 1986, on
that at the detector in the two-slit experiment (in Born’s case, a photo- Bohr’s attitude toward light quanta, or photons).
electric cell), the corpuscular nature of light is revealed simultaneously
According to Bohr, “The tangible content of the idea of light quanta

i
with the wave aspects of the interference pattern, Born proposed the
is limited, rather, to . . . conservation of energy. and momentum” (1929a,
following definition of a particle: “To speak of a particle means noth-
113).* This idea that the particle nature of radiation is essential only in
ing unless at least two pointsof its path are specified experimentally
”
(1969, 101).
17. We see here how arbitrary is the condition of having “two points,” for Born talks
15. Single photon states are unique in this sense: forall of “any point of the path.”If all one needs is two points, this condition is satisfied by the
other states (classical or non- definite position of a light emitter, or electron gun, and of a slit, unless one assumes a
classical), “the probability of a double detection is different
from zero even when the strange transmutability in the same experimental arrangement of electrons or photons,
average numberof photons. .. is less than unity” (Ghose and
Home 1992, 1436). which left the emitter as waves and passed throughtheslits as particles.
16. These authors argue that in interference-type experiments
(Bohrian two-slit experi- 18. Born used this definition while discussing the Comptoneffect (rather than the pre-
ments, for example), the quantum formalism implies the disappea
rance ofthe interfer- vious case of a two-slit experiment): “The corpuscular description means at bottom that
ence pattern whenever onetries to obtain “which path”(particle)
information, thus sup- we carry out the measurements with the object of getting exact information about mo-
porting the mutual exclusivity of waves and particles.
mentum and energy relations” (1969, 97).
 236 Chapter Eleven
The Myth of Wave-Particle Complementarity 237
the interaction of radiation and matter wasat
the core of Bohr’s idea of
the mutual exclusivity of wave and particle ible to each other, as an analysis of the Comolecture
pictures: “It is not possible shows. In the
in a single picture to accountfor the various Comolecture, as I have argued, “complementarity between
properties of radiation. For space-time
certain properties the idea of wave Propagatio and causality” is an umbrella concept covering several basica
n is quite essential. For lly differ-
the study of the energy and momentum transferre ent physical situations. In elucidating the applicability of conser
d in the indiv idual vation
processes, the photon idea, close to the partic lawsin the Bothe-Geiger experiment, Bohr applied causalityto particl
le concept, is equally es- es,
sentia l” (1957b [lecture 3], 609). . and space-time notions to waves. In discussing the limits of space-t
: ime
As I have previously argued, Bohrdid not assig models for an atom in a definite stationary state, he correlated space-
n arealistic signifi- time with particle models, and causality with the notion of a wave (a
cance either to the wave picture of matter or
to the particle picture. — ‘definite stationary state is represented by a constant value of energy
of radiation (a point also made by Murdoch 1987,
78). These pictures, _ and by a single harmonic wave). The frequent assumptionthat one kind
as opposedto realistic models of waves and particles
in the classical of complementarity reduces to the other is unfounded and leads to
realm, are needed for “visualization,” for
adapting oneself intuitively
to a nonintuitive quantum world. This confusion.
being the case, it is not clear
why,
and in what sense, they are “equally necessary”
(Murdoch 1987, 78). Ideological and Pedagogical Uses of Wave-Particle Complementarity
Noris it clear, one might add, whyit is important that
they be mutually
exclusive: why should imprecise and limited analo
gies be consistent If Bohr’s wave-particle complementarity is neither unambiguous nor
with each other?
Bohr’s idea of the mutual exclusiveness of waves and necessary in theoretical research, whatis its function? To what usesis it
particles hadits put and whataims doesit serve?
roots in the Comolecture. Whatheconsidered contradictor
y; and there- Orthodox visualizable discussions of wave-particle complementarity
fore mutually exclusive initially, were the
ideas of a single, infinitely
extended harmonic wave and a precisely localiz have clear ideological and pedagogical objectives. This is the main rea-
ed, free material par- son quantum theorists, most notably Heisenberg and Born, sometimes
ticle (chapter 6). Yet their contradictory natures did not
bother Bohr— supported Bohr’s discussions of wave-particle complementarity in pub-
they were “abstractions,” unrealizable in physical
experimental sit- lic. Their support for an accessible, Bohrian interpretation becamees-
uations. The contradiction’ between these two
ideas, Bohr argued, is
resolved by the de Broglie~Schrédinger wave packet. | pecially strong when they targeted their expositions toward wider au-.
As Bohr’s emphasis on the usefulness of wave image diences and not only the mathematically initiated: “The existenceofthis
ry wasreplaced mathematical theory [quantum mechanics] showsthat the whole struc-
gradually by an operational emphasis on measuremen
t, his method of tureis logically coherent. Butthis proofis rather indirect and convinc-
avoiding contradictions changed: moving away
from the notion of a ing only for those who understand the mathematical formalism.It is
wave packet, he turned to discussions of the mutua
l exclusivity of ex- therefore an urgent task to show directly for a number of important
perimental arrangements for disclosing wave and
particle properties, cases why, in spite of the use of two such different pictures as particles
respectively. Bohr adopted wave-particle complementar
ity by analog
y and waves, a contradiction can neverarise” (Born 1936, 47).
with kinematic-dynamic (space-time and causali
ty) complementarity. The status of wave-particle complementarity is closely tied to the
In the case of kinematic-dynamic complementarity,
which followed in Bohrian doctrine of the necessity of classical concepts. If one gives up
a straightforward way from Heisenberg’s uncert
aintyrelations, the mu-
tual exclusivity of the respective experimental this controversial, if not unfounded, doctrine, the basic role of wave-
arrangements was de- particle complementarity fades away. The necessity of classical con-
manded for the consistency of the quantum mechan
ical scheme. cepts, the overthrow of the concept of reality (the impossibility of a
This misleading analogy between wave-particl
e and kinematic-
dynamic complementarity resulted in the wides quantum ontology), and the finality of indeterminism— all are tied to
pread misconception
that there mustbe a precise connection between wave-particle complementarity in the overall rhetoric of the inevita-
the two. Not surpris- bility of the Copenhagen orthodoxy. Not surprisingly, those physicists
ingly, different commentators interpreted this connec
tion in different who opted for someversion of realism—Bohm, Schrédinger, Einstein,
terms—some correlated energy and momentum with
particles, and Landé— rejected Bohrian complementarity. Wave-particle duality is also
Space-time with waves (Born 1969, 97), while
others identified particles
with space-time. Yet the two kinds of complementar extensively used in the rhetoric of the inevitability of indeterminism.
ity are not reduc- Born’s assertionis typical: “It is clear that the dualism, wave-corpuscle,
 238 Chapter Eleven
The Myth of Wave-Particle Complementarity 239
BIRR OMECE AUREE

and the indeterminateness essentially involved therein, compels wave-particle duality expressed in the idea of a wave packetis not com-
us to
abandon any attempt to set up a deterministic theory” (1969, 102,
m patible with Bohr’slater elaborations of complementarity, which empha-
siontr eta -ahnnak So

italics). The philosopher Norwood Hanson (1958, 1959) followed

the size the mutual exclusiveness of wave andparticle attributes. The con-
orthodox guidelines in arguing that wave-particle complementarit
y tinuing confusion between wave-particle duality and wave-particle
necessarily excludes a deterministic alternative to the Copenhagenin- contplementarity gives the impressionthat most physicists accept wave-
terpretation (for a full discussion, see chapter 14). Wheaton openedh
is particle complementarity, while they only endorse wave-particle du-
valuable historical description of the wave-particle issue in the early ality. Born and other physicists sometimes followed Bohr in his rhetoric
days of quantum theory by emphasizingits importance to the “over-
of the inevitability of wave-particle duality: “The lasting result of Bohr’s ©
throw” of determinism: “Thefirst years of this century witnessed the
endeavorsis the simple consideration given above[analysis of the wave
final rejection of determinism in physical theory; there is no more com-
packet] which showswith irrefutablelogic that the Planck—de Broglie
pelling example of this than the synthesis forgedin the early 1920s be-
laws of necessity implythe duality[of] particles-waves” (1953a, 128).
tween theories of matter and theories of light” (1983, 3).
As already noted, the Gottingen-Copenhagen physicists later as-
The most extensive use of wave-particle duality was the simple deri-
cribed a realistic wave meaningto a single particle (a three-dimensional
vation and legitimation of the uncertaintyrelations.” The easiest,
most wave packet) but continued to treat the y-function for higher dimen-
accessible way to demonstrate the uncertaintyrelations is byident
ify- sions as an abstract, purely mathematical concept. This artificial divide
ing a particle with a limited wavefield, the argument Bohr presented between tnree and more dimensions has importantimplications. If
in the Comolecture, and consequently the way many textbooks intro-
wave-particle complementarity applies only to free particles, the con-
duce the uncertainty formula. Heisenberg, who considered the uncer-
ddwenvete

cept is too limited to be of general significance and wave-particle com-

tainty relations fundamental to arguments for the self-consistency
of plementarity becomes marginal, if not superfluous. If, for reasons of
quantum theory, was willing to use Bohr’s accessible dualistic explana- homogeneity and consistency, one applies the abstract probabilistic
tion of the uncertainty relations for wider audiences and for pedagogi- particle interpretation to a single particle as well, then the wave packet
cal purposes. As helater revealed, he used Bohr’s complementarity be- in the three-dimensional case is as “abstract” as in the multidimen-
cause “it did not do any harm”to his ownexplanation, yet he did not sional case and simply signifies the probability of a particle’s having a
believe “it was necessary”(interview with Heisenberg, AHQP).
certain position value. © se _*
The reason wave-particle complementarity is not necessary for
an If one extends the basic particle substratum, guided by probabilistic
elucidation of the uncertainty relations is that Heisenberg’s formula
fol- lawfulness, to the case of a single particle, the whole argument for
lows from the mathematical formalism of Dirac-Jordan transformation wave-particle complementarity dissolves. We do not have in this case
theory, while wave-particle complementarity does not. As Jammer
put reality of matter waves even for a single particle, viewing it instead
it: “Complementarity is an extraneousinterpretive addition toit [to
the as the three-dimensional case of the same probability calculus that is
formalism]” (1974, 60-61). Moreover, wave-particle complementarit
y used in higher dimensions. The last stand is implied in the following
is always discussed for a very limited, physically uninteresting case—
remark by Feynman:“Although one may betemptedto think in terms
that of a free particle. Bohr elaboratedthe de Broglie-Schrédingeridea
of ‘particle-waves’ when dealing with one particle, it is not a good
of a wave packet and the Planck—Einstein—de Broglie relation
s (E = hv, idea. . . . For if there are two particles, the amplitude [probability] to
p = hk), which Schrédinger had used to argue the inapplicability
of find one at 7, and the otherat 7, is not a simple wavein 3-dimensional
particle concepts in the microdomain, into an argument
for wave- space, but depends on six space variables 7, and 7,” (Feynman, Leigh-
particle complementarity, uncertainty, and indeterminism.2°
In fact, the ton, and Sands 1969, 3:3—4). This point of view is indeed consistent
with Born, Heisenberg, Pauli, and Jordan’s particle-kinematic probabi-
19. Conversely, the use of wave and particle visualization
mula must be taken into consideration: One can
means the uncertainty
for- listic interpretation. Yet, following Bohr, the orthodox physicists often
apply the wave or particle picture by
es into consideration limitations imposed by the uncertai invokedthe three-dimensional wave packet as more than a formal anal-
nty relation (Heisenberg
58, 43). ogy. Pauli, in popular presentations, usedthe idea of a wave packet not
20. Initially, Bohr used the Planck-Einstein formula E merely as a visualizableillustration but as a physical explanation for
= hy to argue the inapplicability
of particle concepts—see his Bohr-Kramers-Slater theory, the uncertainty relations: the uncertainty relations are valid because in
in which, due to the “irratio-
nality” of this formula, Bohr was eagerto avoid the
at any price.
corpuscular conceptof light quanta wave imagery no packets exist that contradictthe relation AxA(1/A) ~ 1
 240 Chapter Eleven
iR se SS

The Myth of Wave-Particle Complementarity 241

(the relation between the spatial limitation of a group of
superposed ation” (1956, 133). Born and Walter Biem (1968) published a highly criti-
waves and wavelength A; Pauli 1950).
If wave-particle complementarity is an arbitrary additi
cal rebuttal of Landé’s views, using the characteristic rhetorical tactics
on to quan- of discrediting the opponent and appealing to authority. Landé “does
tum theory, wecan easily understand whyit generated so muchst
erile not realize the historical origin of the dualistic interpretation” and does
verbiage.” Born again appears on the scene with a rhetorical
— ———————rnrneeamnmntnnnentseati taniS

sword in not * correctly describe its physical meaning.” Heis driven by “preju-
hand (see my previous discussion of Born’s confrontation with
Schré- dice” and “dogma,” and he “ignores important physical discoveries”
dinger on quantum jumps), this time instigated by Landé’
s (1965) criti- (1968, 51). Dualism is a “discovery, and not an invention” (1968, 54) of
cism of wave-particle complementarity. Landé argued that “duali
ty” Einstein himself, made when Einstein was young and drew “irrefutable
might sometimes be “helpful for heuristic reasons” but thatit is totally
~ inferences” (1968, 51), as opposed to Einstein‘s deterioration into spec-
“unphysical.” He stressed eloquently and repeatedly that the “crucia
l” ulation in later years. While every physicist must accept Duane’srule,
evidence for matter waves, supposedly impossible to explai
n other- which “describes correctly all experiments of momentum exchange” on
wise— diffraction phenomena—is adequately explained
by Duane’s periodic structures (1968, 51), the rule is “obscure” without de Broglie’s
“third rule” of the quantizationof linear momentum for a
diffractinglat- idea of the connection between particles and matter waves. Besides, the
tice, whichis a bodyperiodicin certain space directions and which there-
dualism Landé himself introduces (“particulate nature of matter and
fore can change its momentum component only by discrete amounts.
wave nature of light’) is even more unsatisfactory. And while Landé’s
This conditionyields a set of discrete angles of electron deflection that
" derivation ‘of the quantum mechanical probabilistic rules “can be inter-
determine and explain the diffraction pattern without the assumption
esting in itself,” there is no need, Born and Biem reprimanded,to ac-
of de Broglie waves. Similar considerations can, according to Duane,
companyit “by attacks on supposed enemies”(1968, 55).
explain all of the other “mysteries”thatallegedly can only be accounted
Landé’s (1968) answer wasa fitting rhetorical counterattack: wave-
for bythe “fiction” of the dual nature of matter (Landé 1965, 123).
particle duality “received its death-knell” from none other than Born
- Landé (1965) also did not accept Heisenberg’s conciliatory formula-
himself, through his “admirable”statistical interpretation of the y-
tion of the “symmetry,” rather than the “duality,” of waves and parti-
function in term ofparticles; to present Einstein “as a champion of du-
cles, based_on second quantization:a photon whoseroleis merelythat
alism is utterly unhistorical’; Landé’s own views are no longer those
of a quantum numberattached to the periodic components of the
“of a lonely Don,” but are shared “by many prominent physicists and
continuous Maxwellfield hardly deservesto be regarded as a particle
. philosophers of science” (1968, 56). In the end, Born and Biem had no
Similarly, the fact that atomic probabilitiesfor the distribution of par-
choice but to retreat: “Clearly, it is possible to formulate a quantum
ticles obey a rule of wavelikeinterference rather than of simple (classi-
cal) addition is not a reason to conclude that particles have a “wave mechanics of particles avoiding all wave-like terms.” Still, they asked,
“Whythe effort?” (1968, 56).
nature,” but rather a stimulusto search for a simple and logical foun-
dation from which such interference (or equivalent rules for matrix
multiplication) for probability amplitudes can be deduced. Such
a
simple and logical foundation, fromwhich Landé succeeded in deduc-
ing the interference of probabilities, contains the non-quantum postu-
lates of symmetry, order, and coherence (Landé 1955).
‘
Landé’s derivation of the interference of probabilities was largely
ig-
nored by representatives of the G6ttingen-Copenhagenallianc
e, but
his attack on the Copenhagen credo wascountered at once. Rosenf
eld
dismissed Landé’s work as “making a muddle of a perfectly clear situ-
21. In contrast to the idea of a wave-particle synthesis, which, I argued
earlier, served
as a fruitful heuristic principle.
22. The basic innovation in the traditional view in physics and philosop
hy, according
to Landé, was not “wave-particle duality” or “reality” but
the irreducible statistical law-
fulness in the quantum domain. i
 CHAPTER 12
EF

Complementarity as Metaphor

Perhaps our ability to convince others dependson theintensity with which we

can persuade ourselvesof the force of our own imagination.
Wordsattributed to Niels Bohr by Heisenberg1971; 131

Wehaveat the end only to take recourse to painting with wordsjust like one -
paints, as [an] artist paints with colors just trying to use them in such a
way [as] to be able to give to one anotherjust an impression of certain
connections of certain harmonies.
Niels Bohr 1937d;354 ~
Bohrfelt that whenever one came with a definite statement about anything,
_ one was betraying complementarity. ;
Interview with Leon Rosenfeld, 22 July 1963, AHQP

Introduction

Bohr was an avid storyteller. A whole generation of physicists was

raised on Bohr’s stories. As Pais reminiscences: “Sooneror later, for
“4 the purposeofillustrating some point..... Bohr wouldtell one or more
stories” (1991, 6). The most inspiring, never-ending story, which Bohr
nevertired of repeating, was aboutthe “great interconnections”in sci-
ence andlife as revealed by complementarity. But what is complemen-
tarity? It is not a principle, as both Bohr and Heisenbergoftenstressed.
Even less is complementarity a model. A model of quantum theory
should contain something essentially quantum. Complementarity does
not. Complementarity is a metaphor, powerful enoughto cut across
many domains, inspiring enoughto construct a new sensibility.
Complementarity is not a rigorous guideto the heartof the quantum
mystery. Nor do Bohr’s numerous analogies between quantum physics
and other domains, such as psychology or biology, withstand close
 244 Chapter Twelve
Complementarity as Metaphor 245
scrutiny. Complementarity does not reveal preexisting
similarities;it will remain except a fancy formulation of the uncertainty principle.
generates them.! Complementarity builds new worlds by
making new Bohr certainly intended complementarity to mean morethanthat.
sets of associations. These worlds are spiritual and poetic,
not physi- The ambiguity, opaqueness, and obscurity of Bohr’s writings is leg-
cal. Complementarity did not result in any new physic
al discovery — endary. Bohr himself, despite the heavy rhetoric of “unambiguity of
“itis merely a way to talk aboutthe discoveries that have alread
y been

te
made”(interview with Dirac, AHOQP).
communication” that underlay the inevitability of complementarity,
Only great poets and great prophets succeed in impos considered obscurity a virtue.‘ Bohr’s ambiguity was enveloped, by
ing their pri- Bohr and others, in a fog of profundity. “Deep truth” cannot be ade-
vate associations on the whole culture. Einstein considered
Bohr “a quately expressed. It is almost as hard to write about complementarity
prophet,” and Freud the author of a “huge mythology.” Einste
in had as “about religion . . . being almost as long aslife itself” (quoted in
little sympathy for either. Neither Freud nor Bohr wasintimida
ted by Blaedel 1988, 27). Presenting complementarity to a French audience,
Einstein's criticism. In a shrewd reply to Einstein, Freud was
anything Bohr announced that “the situation is very difficult to express in words
but apologetic. Freud had nointention of demarcating psycho
analysis. of any language” (notes for talk given atthe Institut d’Henri Poincaré,
from science: “But does not every natural science lead intima
tely to 18 January 1937, AHQP). :
this—a sort of mythology?Is it otherwise today with your physica! sci- Bohr extended complementarity to psychology, biology, and the
ence?” (quoted in Erikson 1982, 168). , % theory of culture. In this he followed his teacher Harald Heftding,
As is suitable for a prophet, Bohr-talked in fables and parables. Of- “who 'perceived many analogies between physics and psychology (Faye
fending somesensibilities, Heilbron compared Bohr to a”guru” (Heil- 1991). Bohr elaborated Hoffding’s ideas, construing analogies between
bron 1987, 221). As gurus do,in his later years Bohr inspired by per- “disturbance”in introspection and in physicalinteraction.Initially, this
sonal contact. As Otto Frish recalled, “after dinner, we sat close to
was no more than a hunch—Bohrspokehesitantly of “suggestive anal-
him—someofusliterally at his feet, on the floor—so as not to miss a
ogy” (1929b, 100), or of “moreor lessfitting analogies” (101). Ina letter
word” (French and Kennedy 1985, 353). Bohr had the exclusive author-
to Heffding, Bohr revealed that he had “the vagueidea that there might
" ity to reveal the “hidden harmonies”in nature. When one did not un-
be a possibility of proving a similar complementary relation between
derstandBohr, the reaction was not inquiring criticism—it wasa
feel- those aspects of the description of the individual psychological pro-
ing of awe for the “deep and subtle” philosophy of Bohr. Bohr cesses which relate to the emotions and those which relate to the will
expected the new framework of complementarity he had built to be
as that which quantum theory has shown to obtain, with respect to
taught to children in schools,just as the heliocentric theoryis taught
elementary processes in physics” (Bohr to Heffding, 1 August 1928,
(interview with Bohr, AHQP). Even Bohr’s close collaborator and
co- quoted in Faye 1991, 58-59, myitalics). Many of Bohr’s analogies did
author of complementarity, Pauli, thoughtthe “imperialism of comple- not undergo any substantive change overthe years. Rather, what was
mentarity” was going too far. Objectivity,oratleast the illusion thereo
f, suggestive and vague became, merely by virtue of repetition, rigorous
demands some intersubjective agreement. No metaphor can form a ba-
and compelling: “It is not a question of weakparallels. It is a question
sis for “unambiguous communication,” to use Bohr’s expres
sion. Am- of investigating as accurately as we can the conditions for the use of our
biguity is a necessary part of a metaphor's suggestiveness. Strip
com- words” (Bohr 1958e, 704).
plementarity of its imaginative, imprecise associations and not
much When Bohrfirst explored the similarities between physics and psy-
chology, he felt that we “can hardly escape the conviction that we
1. Bohr’s disciple Kalckar wrote: The connection between Bohr’s work - have acquired a meansof elucidating general philosophical problems”
and the whole
of his personality is so close that one can almost speak of an
identity. The turn he gave to
the trend of modern physics and through whichit receivedits far reaching (1929b, 101). It is this conviction that sustained repetition without sub-
epistemologi- stantive elaboration. The initially novel and fresh metaphors became
cal consequences, arose so directly from and harmonized in such a rare
degree in hisown
mind, that one dares to use of him the phrase which one would
otherwise reserve for the familiar, obvious. ““Vague” analogies become “beautiful examples,”
greatartists: that he created a world from within” (1967, 229). “striking analogies” (Bohr 1958e, 704). .
2. “Only those who knewhim personally could experience the
immense inspiration According to Heisenberg, Bohr was a “natural philosopher,” not “a
exuding from his intuitive grasp of physics ‘and his humane personali
ty” (Pais 1991, 29).
3. Using the expression “deep and subtle” has become almost obligator
y when writing 4. Bohr especially liked a fable about the talk by a rabbi that was understood neither
about Bohr. : by his listeners nor by the rabbi himself (Pais 1991).
 246 Chapter Twelve
Complementarity as Metaphor 247
mathematical physicist.” For a mathematical physicist,
to understand nature of matter, or betweenthe unity of personality and the individu-
or explain means to construct a suitable set of model
s and to specify the ality of material particles— contain no more objective content than the
rules for connecting the abstract theoreticalentities
with empirical data. correspondence between dreams and winning lottery numbers outlined
Mathematical physicists explicate the meaning
of operators, eigenfunc- in an essay by his grandfather.>
tions, observables, projections in Hilbert spaces
. Bohr had no use for Many of Bohr’s analogies that initially seem appealing fall apart
any of these notions. For Bohr,as for 2 pre-Newtonia
nnatural philos
o- when probed more closely. In his early writings, Bohr often draws a
pher, “any explanation or analysis only means
to use analogies from parallel between atomic measurementandintrospection.In bothcases,
simpl e experience” (Bohr 1937d [lecture 6], 353). Vico’s
description of apparently, it is inherently impossible to separate the observing from
whatis distinctive about proper philosophizing
fits Bohr especially the observed. Bohr’s conclusion wasthat atomic interaction is in-prin-
well: “Specifically philosophic quality .. . [is the]
capacity to perceive ciple “unanalyzable” and “unsurveyable”’—a prohibition not -many
analogies existi
ng between matters lying far apart... . It is
which constitutes the source and principleof all
this capacity working theoretical physicists would choose to comply with. Manyof
ingenious, acute and Bohr’s examples of mutual exclusivity are far from convincing. The of-
brilliant forms of expression” (quoted in McMullin 1991,
57). ten repeated example of complementarity between “thoughts” and -
Bohr was a philosopher of “harmonies,” of symbo
lic meanings. As “feelings” is contradicted by contemporary psychological research (see
Pauli did not fail to notice, Bohr’s correspondence betwe
en macro and Lazarus, Kanner, and Folkman 1980 on the connection between emo-
micro, based on an analogy between a planet
ary system and a micro- tion and cognition). The mutual exclusivity of reason (associated with
scopic atom, retained the original medieval notio
n of harmonies be- the masculine) and emotions (associated with the feminine) has also
tween macrocosm and microcosm. Yet, if the corres
pondenceprinciple, been subjected to penetrating critique, by feminist scholarship. Though
by an ingenious mathematical elaboration, especially
in the hands of some of Bohr’s associations remain private and inexplicable, others,
Kramers and Heisenberg, led to the new quantum theory
, complemen- such as ‘the preceding example, resonated within the culture of his
tarity neverleft “common language.” By not leavin
g common language,. time. Heisenberg, who joined Bohrin the dissemination of his meta-
Bohr could noteven attemptto construct a quantum
ontology. Bohr and phors, understood well their persuasive power. Thoughto his col-
his followers presented this idiosyncratic choice
as a prohibition in leagues Heisenberg preferred to talk in unambiguous mathematical
principle. Not surprisingly, many of Bohr’s thoug
ht experiments in- language, he appealed to.the general ‘mental climate” whenlectur-
tended to demonstrate complementarity contain
nothing quantum in ing to popular audiences.® Bohr’s numerous “harmonies” werethose
their analysis (Beller and Fine 1994). °
If there is a single idea that inspires Bohr’s analogies,i “useful (sinnvoll) interrelations ‘belonging together’ within the human
t is the break- mind” (Heisenberg 1979, 68).”
downofthe classical concept of motion, andthe
loss of continuity (in- Bohr’s writings on complementarity are metaphorical in a strong
finite divisib
ility) and of the visualizability (Anschaulichkeit)
associated sense. Complementarity is an ‘artificial word,” which “serves only
with it. It is this idea that inspired, often mustakenly,
Bohr’s intuitive briefly to remind us of the epistemologicalsituation” in quantum phys- ©
discussion of the interconnection between reality,
acausality, and loss ics (1937c, 293). Because the situation is unprecedented, one can only
of visualizability in the quantum domain.This idea
also inspired Bohr turn to “quite other branches ofscience, such as psychology or. . . to
to see a general epistemological similarity betwe
en atomic physics and thinkers like Buddha and Lao Tse” (1937b,20). But these disciplinesare
psychology: “We maysay that the trend of moder
n psychology can be themselves ambiguous and in need of morerigorous explication, so one
characterized as a reaction against the attempt
at analyzing psychical has to turn back to quantum physics, where matters can be explicated,
experienceinto elements which can beassociate
d in the same wayas are so Bohrintimated, more precisely. Clearly, this state of affairs cannot be
the results of measurementsin classical physics”
(1938b, 27). The loss of understood by any “substitutive view” of metaphor, in whichtheliteral
visualizability in physics inspired many
ofBohr’s far-fetched analogies:
“emotions and volitions” are similar to the
quantum of action because 5. Pais described Bohr’s grandfather's “witty essay” in which “the dream table estab-
they are incapableof being represented byvi
sualizable pictures. Impro- lishes a correspondence between a specific type of dream and the lottery numberto be
vising on the complementarity theme, Bohr picked” (1991, 35).
gavefreerein to his imagi-
nation. Many of his analogies — for example, 6. “Every scientific theory arises in a certain mentalclimate . .. the authorofthe theory
between the quantum of
action and the conceptof life, between’the flow may be only vaguely conscious of it” (Heisenberg 1979, 65-66).
of thinking and the wave 7. 1 would translate sinnvoll as “meaningful,” rather than “useful.”
 248 Chapter Twelve
Complementarity as Metaphor 249
meaning of the primary subject can replace the metaphorical meanin
g “emphasis on permanent recording under well-defined experimental
of the secondary one. Even the “interactive” view of Max Black (1962)
conditions . . . corresponds to the presupposition, implicit in the classical
is only oflimited help. According to this view the primary and second-
physical account, that every step of the causal sequence of events in
ary domains (for example, physics and psychology) interact
so as to principle allowsof verification”(myitalics). It is this “correspondence”
produce novel understandings of both, but Black assumes that the pri-
mary domain is understood to a considerable degree. No such well- that*allows one to view quantum mechanicsas fulfilling “all demarids
understood primary domain exists in Bohr’s analogies. Rather,
on rational explanation with respect to consistency and completeness”
we en-- (Bohr 1958c, 6). By presenting numerous “correspondences” between
counter here a hermeneuticalcircle. Bohr’s complementarity should be
approached as a foreign culture, as a.newly encountered, unfamiliar the formalism and the “conditions of experience,” Bohrcreated the il-
" lusion that the quantum formalism is a direct confirming instance of
symbolic world. Clifford Geertz’s characterization of the anthropologi-
cal approach is a fitting one to describe the encounter with the strange complementarity, thus enforcing the acceptance of an unfamiliar for-
framework of complementarity: “Hopping back and forth betwee malism and a newphilosophy of physics concurrently.
n the Heisenberg, when presenting the lessons of quantum theory to wider
whole conceived throughthe parts which actualize it and the parts con-
ceived through the whole which motivates them, we seek to turn them audiences, sometimes adopted Bohr’s language of correspondences.
Arguing that the quantum formalism represents our knowledge rather
by a sort of intellectual perpetual motion into explications of one an-
than the objective course of events, Heisenberg pointed out the corre-
other” (1976, 236). ae ee ‘lation, or reciprocal “image,” betweer the act of registration, in ob-
In this hermeneutical web of metaphors, “hopping back and forth”
servation and the collapse of the wave function: “The discontinuous
often interchanges what is foundational and what is derivative, ex-
change in the probability function, however, takesplace with the act of
planandum and explanans. As I will shortly argue, Bohr’s language of
registration, because it is the discontinuous change in our knowledge
“correspondences,” conjoining certain concepts and thus creating “har-
monies,” was inspired by powerful, thoughoften misleading, images. in the instant of registration that has its image in the discontinuous
change in the probability function” (1958, 54, myitalics). Earlier, Born
His understanding wasthat of a poet, a mystic, an.artist—Bohr’s own
_, Words (this chapter’s epigraph)characterize his thought especially well: had advancedhis statistical interpretation of the wavefunction and his
statistical solution of the collision problem also byasserting a preestab-
“We have at thé end only to take recourse to painting with wordsjust
like one paints, as [an] artist paints with colorsjust trying to use them
lished harmony between the possibilities of theorizing and the possi-
bilities of experimentation (Beller 1990).
in such as way[as] to be able to give one another just an impression of
Bohr’s writings are permeated not only with harmonies between
certain connections of certain harmonies” (1937d, 354).
theoretical advances and experimental possibilities but with what he
considered necessary connections between key concepts, such as cau-
The Web of Correspondences and Harmonies
sality, visualizability, objectivity, and the distinction between subject
and object. These connections, firmly entrenched in Kantian philoso-
Bohr’s thought teems with “correspondences” and “harmonies.” If, in
the old quantum theory, such correspondences served as powerful heu- phy andin the classical idea of motion, Bohr fully embraced. His phi-
tistics (Jammer 1966; Darrigol 1992a), after 1927 Bohr advanced losophy of complementarity, as I have argued, sprang from his basic
the idea intuition of the breakdownof the classical idea of motion (overthrow
of correspondences, or harmonies, to legitimate an unfamiliar and
ab- of visualization, causality, and reality) and wasaidedinitially by the
stract quantum theory. The idea of harmonybetweenthe possibilities
of observation and theoretical definition pervades, as [have argued, the incorrect idea of disturbance (the impossibility of demarcating object
Comolecture. Later, when Bohr elaborated thelessons of complemen- from subject, the inseparability of phenomenaand their means of ob-
servation, and the subsequent “indivisibility,” or “wholeness,” of the
tarity, he argued that features of the quantum formalism “correspond”
to the freedom to choose the appropriate measuring devices.’ Similarly, quantum interaction; chapter 7). In the classical idea of motion, space-
time concepts, conservation laws, and visualizability (Anschaulichkeit)
8. “The possibility of disposing of the parameters defining the quantum are intimately connected: from given initial space-time conditions and
mechanical
problem just corresponds to our freedom of constructing and handling the conservation laws (causality), one can calculate the subsequent dy-
the measuring
apparatus, which in turn meansthe freedom to choose betweenthe different
tary types of phenomena” (Bohr 1948, 452).
complemen- namical evolution of a system, as well as form a model (a picture, a
visualization) of its behavior in space-time. Bohr’s initial conflation of
 R
a SERRES|TH
250 Chapter Twelve
Complementarityas Metaphor 231
the ideas of causality and reality, and of causality
and visualization, postulate, rather than explain, the inseparability between measuring
hadits sourcein the classical idea of motion—Bohr thought intuiti
vely and measured, as well as the indivisibility, or wholeness, of quantum
that the breakdownof the classical idea of particle motion implied
the phenomena." Bohr’s wholeness often meant nothing more than an
breakdown ofboth causality and our concept ofreality.’
Bohrcharacterized classical description as “causal pictori operational definition, where the meaning of a conceptis intrinsically
esate

al” (1960c, tied’to the procedure of measurement. Howis then the “nonpictorial”
11), and he retained this intuitive, yet incorrect, idea of
a necessary
IO)

connection between causalityand visualizability throughouthis character of quantum physics suited to embracethe indivisibility of
life, quantum phenomena? If the phenomenaare defined relationally, or -
even as his ideas about reality and acausality change
d (chapter 7). If contextually, thereis little reason for discarding causality a priori. Not
visualizable classical physics is “causal,” then it is natural
that the surprisingly, in Bohm’s version of quantum mechanics wholeness does .
, honvisualizable quantum physics be noncausal: “In
conformity with notlead to acausality. In fact, Bohr retainedthe intuition of a correspon-
the non-pictorial character of the formalism,its physicalinter
pretation dence between wholeness and acausality from his older, discredited
finds expression in laws of essentially statistical type”
(Bohr 1958c, 3v framework of disturbance. _
myitalics). us
Bohr’s metaphorical framework of harmonies and corresponden Thus Bohr’s complementarity is built on a clustering of associations,
ces which are nowhere grounded. Bohr’'s liberal interchangingof explanan-
is sustained by such expressions as “finds logical express
ion,” “is dum and explanansleadsto his peculiar assertion that Heisenberg’s un-
i suited,” “it is not surprising that.” Consider two characterist
icexamples: ' certainty relation follows fromcomplementarity: “The ultimate reason ©
“Since, ... we cannot neglect the interaction between
ii
the object and the that in no conceivable measurement conjugate quantities can befixed
instrumentof observation . . . , if is not surprising that in
i all rational with a greater accuracy than that given by(5) [the uncertainty formu-
applications of the quantum theory, we have been concerned with es-
sentially statistical problems” (Bohr 1929b, 93, myitalics). And many las] is indeed the complementary character of the pictures employed”
years later: “We are dealing [in quantum mechanics] witha mathe- (1939, 391). Complementarity, begot by repetition, assumesa life ofits
ne mpepppeesca

matical generalization of classical physical theories which by.its own and begins to serve as a basic explanatory concept, from which
non- other aspects of the Copenhagen philosophy seem “inevitably”to fol-
pictorial character is suited to embrace the indivisibility
of quantum low. When he was in a Copenhagen mood, Heisenbergtoo freely in-
phenomena”(Bohr 1957a, 669, myitalics).
The two examples sound innocently similar. Yet they are worlds terchanged explanandum and explanans. He often argued, using dis-
apart. Thefirst is taken from Bohr’s early writings at the time whenal turbance imagery, that the atomic constitution of matter leads to the
l uncontrollability of the measurement interaction. Yet sometimeshis
of his considerations were informed by the idea of disturbance.
It is reasoning was reversed—it is the uncontrollability of the interaction
with the help ofthis idea that Bohrjustified the connection betwee
n the that necessitates thefinite divisibility of matter: “The existence ofele-
indivisibility of measurement interaction and the necessary
overthrow mentary particles is only justified by this fact [uncontrollability ofinter-
of causality and reality (chapters 7 and 9). As Bohr himselflater re-
pudiated the misleading idea of disturbance, nointuitive ground action]” (1952, 73). .
eet

re- The Copenhagen frameworkis built on “correspondences” and “har- -

mained for his analysis of the inevitability of indeterminism
and the monies” among a cluster of intuitively interrelated ideas, despite the
overthrowofobjectivity in the quantum domain. Before 1935,
the idea of fact that the meaningof these ideas and the natureoftheir interrelation
disturbance provided theillusion that Bohr’s argument was
grounded vary over time. It is this clustering that misled Bohr’s readers into re-.
in a solid, experimental state of affairs. Noillusion ofthis
sort could be garding his conceptual frameworkasbasically stable. Because his rea-
retained whentheidea of disturbance was discarded
and Bohr declared
that quantum theory was merely a tool for the description soning is essentially metaphorical, Bohr often uses such expressionsas
and predic- “harmonizes, wut corresponds,” eG“is suited to,” or “finds proper expres-
tion of measurementresults. In the post-1935 framework
Bohr had to sion,” rather than “follows from” or “is deducible from.” Herelies a
9. Supplemented by his doctrine of the indispensabili
ty of classical concepts, this in-
tuition led Bohr to deny the Possibility of an objective 11. One could, of course, turn to the quantum formalism and fully acknowledgeits
causal theory, such as Bohm’s.
10. One can have a nonvisualizable deterministic descrip newfeatures of nonseparability and nonlocality. Yet Bohr, who refused or was unable
tion, as well as a visualizable
nondeterministic description. (see the discussion of Bohr's attitude toward mathematics below) to leave the “common
language,” avoided this option altogether.
 oesay
252 Chapter Twelve
Complementarity as Metaphor 253
Teepe
%

striking difference between the analogical use of the correspondence

principle and the metaphorical use of complementarity. The readers of his book, a significance much deeper than what any one of
correspon- us could easily explain” (Bohr 1931a, 177).
denceprinciple guided physicists, by suitable analogies, to explore
the
ay Se: a eee ee

new quantum realm in ways suggested by regularities existing Heffding’s general idea of wholeness permeated Bohr’s writings on
in the complementarity from his Comolecture until the end ofhis life. Like
well-explored (primary) classical domain. No such primary
Sy RReepeaneee

domain the terms “complementarity” and “causality,” the word “wholeness”

exists in the complementarity framework. The complementarit
y prin- in Bohr’s writings has manydifferent meanings and is exemplified by
ciple was a metaphoricaltoolof legitimation—it led tono new
physical different analogies in different contexts. Someof these analogies con-
knowledge. ac : flict with each other, while others are misleading (see the discussion
below). None of these analogies adequatelyreflects the post-Bell no-
“Wholeness” as Metaphor
tions of “inseparability” and “nonlocality.”
“Hunger for wholeness” permeated the early twentieth century’ In whatfollowsI will discuss several of Bohr’s metaphorical uses of
eseagarpInty:?

s Wei- the idea of wholeness. In these discussions Bohr constructed an illu-

mar culture (Gay 1968, 70-101).This hunger was'not confined to the
Weimar republic. In Denmark manyintellectuals inherited the craving sion of explanation, by using outmoded traditional (classical) princi-
for “wholeness,” “unity,” “irrationality,” and “unanalyzability” from ples and ideas. Bohr’s early intuition of wholeness drew strength from
Kierkegaard’s existential philosophy. The young Bohr admired Kierke-- his misleading notion of disturbance. Wholeness, basedon finitude and
gaard (see Holton 1970 and Jammer 1966 on Kierkegaard’s impact on * atomicity, is expressed, according to Bohr, by the prototype of the quan-
Bohr). The wholeness, interconnectedness, and unity of science was, tum jump—’an individual process, incapable of moredetailed descrip-
albeit from a different angle, a central concept in Kantian philosophy, tion, by which the atom goes over from oneso-called stationary state to
which, as a result of the “back to Kant” movement, dominated aca- another” (1929a, 108).
demicphilosophy in German-speaking countries.” As Bohrlater argued, the essential characteristic of the quantum in-
: teractionis its uncontrollability, not its finitude. I have argued that the
' From his neo-Kantian teacher and friend Harald Heffding, Bohr
learned that “true knowledge does not consist of accumulated experi- uncontrollability follows from the mathematical formalism of quantum
re
#He
’ ences butis insight into the interrelationships between experiences” mechanics (chapter 9)—it cannot be deduced from a general philo-
Ie (Faye 1991, 9). In a Kierkegaardian vein, Haffding attempted to create sophical principle. In Bohr’s pre-1935 writing, ““uncontrollability” fig- ;
ures in his discussions of the uncertainty ptinciple—uncontrollability
“unity and harmony of the opposing views” (Faye 1991, 12) and
a ho- prevents a circumvention of uncertainty. While Bohr avoided discuss-
listic notionof science that integrates physics and psychology. Bohr’s
early analogies between quantumphysics and psychology, and his gen- ing the mathematics of quantum mechanics, he appealed to theage-old
eral notion of wholeness, are due to Hoffding.It is from Heffding, ac- metaphorof “hiding” and elusive Nature, who prevents the inquirer
EASOS
cabsRonse

cording to Bohr’s own words,that he learned about the “relativity of from getting too close to her and “penetrating” her secrets. Thus, Bohr
all our concepts” (1932, 200, 203). Hoffding’s impact was one of argued, we cannot know the position of particles in an atom ina given
those
1%

meaningful mental processes that cannot, according to Bohr, be stationary state, because the use of measuring instruments will imply

dute
fully “an exchange of energy between the atom and the instruments which
explained or analyzed: Hoffding’s “influence and guidance could be
24
i

ide
ab- completely hide the energy balance” (1932, 201, myitalics). This idea
sorbed almost without being felt by the one whoreceived it” (1931a,
178). It was because of Heffding’s discussions about the “balanc drew its metaphorical strength from an analogy with biology: “In every
e be-
tweenanalysis and synthesis” that Bohr maintained throughouthis life experimenton living organisms there must remain someuncertainty as
the centrality of the idea of “wholeness”: “While the whole may be built regards the physical conditions to which they are subjected, and the
of individual parts, the appearance ofeach individual partis influen idea suggests itself that the minimal freedom we mustallow the organ-
ced
in turn by the whole.” Haffding’s discussions had for Bohr and ism will be just large enough to permit it, so to say, to hide its ultimate
“for secretsfrom us” (Bohr 1933, 9, myitalics).
many ofthelisteners at his lectures as well as the even more numero
us Initially, disturbance (we erase phenomena whiletrying to observe
12. See Beller (forthcoming) for Kant’s impactonEinstein’s philosop them) was the metaphorical tool used to express wholeness—nature
hy and Chevalley
(1994)for an analysis of the Kantian context of the Bohrian use of the
and “symbol.” Bohr's usage of“wholeness” also belongsto this context.
terms Anschauung hides its detailed working by our interference, though no knowledge
without such interference is possible. Later, Bohr distanced himself
 254 Chapter Twelve
Complementarity as Metaphor 255
from the disturbance notion and introduced a “distinction in
principle Bohr’s conception of wholeness as eradicating phenomenaif we try
between the objects we want to examine, and the
measuring instru- to probeinto them in moredetail goes back to William James’s analysis
ments” (1935b, 218)."3 of mental phenomena. Bohr himself connected this idea with the idea
Yet the idea of disturbance, thoughhelater criticized it, remain
ed his of wholeness expressed by James: “If you have somethings.. . they are
most vivid, most potent metaphorical tool for discussing quantu
m un- so connectedthatif you try to separate them from each other,it just has
certainty and wholeness without recourse to mathematical formula
s. nothing to do with the actual situation” (quoted in Faye 1991, xvii).
This is the reason Bohr and Heisenberg, while denyingits viabilit
y at According to Weizsacker, “in the winter of 1931-1932 Bohr was con-'
times, regressed into using the idea of disturbance.'* Even whenhe
tre- stantly reading James” (1985, 186). Yet this Jamesian idea of wholenessis
jected the notion explicitly, Bohr implicitly retained the intuition ofdis-
poorly suited to a characterization of quantum entanglementor insep-
turbance andofits “hiding” aspect in his later (especially post-19
35) arability. Bohr’s and James’s understanding of wholenessbetterfits the
discussions of individuality or wholeness. This intuition is
apparent in idea of a chemical compound: if we separate its constituent parts (we
Bohr’s notesfor a talk he gave at the Institut d’Henri Poincaré in Paris
can doit), the phenomenon—the compound— disappears. A chemical
in 1937: “Individual phenomenain quite a new sense in physics. When
trying to analyze, phenomena disappear. They appear only
compoundis radically different from its constituting elements —water
under con- is different from hydrogen and oxygen. Such was the understanding of
ditions whereit is impossible to follow their course” (18 January 1937,
wholeness by Vygotsky, who argued for a holistic approach to thought
AHOQP). The example that Bohr provided was the
two-slit experi- and language.'* Jamesian and Bohrian wholeness, while it may capture
ment—if we could determine, by momentum transfer, throug
h which the idea of a classical interaction, is unable to provide insight into a
slit the particles had passed, we would necessarily have “latitude”
in quantum wholenessthat evades classical analogies.
the position.of the measuring diaphragm and would thus exclude the
. Bohr’s wholeness plays an essential role in his metaphorical web of
appearanceof an interferenceeffect. , seemingly interconnected ideas. Intuitively, but wrongly, Bohr argued
This argumentis repeated in an obscure mannerin the Hitchcock
‘for a necessary connection between wholeness and complementarity
lectures (Bohr 1937d, 323) and is much more lucidly elaborated in “Dis-
(the impossibility of a unified ontology) and between wholeness and
cussion with Einstein“ (Bohr 1949, 217). Bohr claimed that
we cannot indeterminism. Thus, in “Mathematics and Natural Philosophy,” Bohr
say what is going on with a particle between its passage through
a slit stated: “The essential indivisibility of proper quantum phenomena
andits arrival at a detector; measurement doesnot allow
it because un- finds logical expression in the circumstance that any attempt at a well de-
der measurement the “phenomenon. . . disappears entirely” (1937d,
fined subdivision would require a change in the experimental arrange-
323).This reasoning makes sense in an antirealist approach,
where ment that precludes the appearance of the phenomenonitself. Under
it is meaningless to discuss behavior and properties independentl
y of these conditions,if is not surprising [again we meetthe language of cor-
measurement—here “analyze” means “to measure”: “we actually can-
respondences] that phenomena observed with different experimental
notatall analyze in such an arrangement. . . whatis going on from
the arrangements appearto be contradictory whenit is attempted to com-
time the particle comesin untilit is caught. We cannot control that
with- bine them in a single picture. Such phenomena may appropriately be
out wiping the phenomenon out entirely. The phenomenoni
s in that ~termed complementary” (1956, 559, my italics). And on the existence
Sense an individual phenomenon, just like the individual
process of of a “natural” connection between wholeness and determinism Bohr
transitions between thestates of the atom” (Bohr 19374,
324). wrote: “The indivisibility of quantum phenomenafindsits consequent ex-
pression [!] in the circumstance that every definable subdivision would
13. Note how different the idea of “irrational,”
‘ ‘hiding” Nature is from Einstein’s ar- require a change in the experimental arrangement. . . . Thus, the very
gumentfor the comprehensibility of nature: his belief that the “subtle,
God would not construct a world that fundamentall but not malicious”
y hides part of its workings.
14. The.electron has been pushedby the light quantu 16. “Two essentially different modes of analysis are possible in the study of psycho-
m,it has changed its momentum logical structures. . . . The first method analyzes complex psychological wholesinto ele-
andits velocity, and one can showthat the uncertainty
of this changeis just big enough to ments. It may be comparedto the chemical analysis of water into hydrogen and oxygen,
guarantee the validity of the uncertainty relations
(Heisenberg 1958,
47-48, myitalics).
tunmsae is notthat the “phenomenon disapp neither of which possessesthe properties of the whole. . . . Psychology. . . analyzes verbal
ears” butthat the laws of quan- thought into its components, thought and word, and studies them inisolation. In the
oe oe nics do notallowone to obtain the specifics
of the previoussituation bycal-
ulation. Again argumentsfor consistency are course ofanalysis, the original properties of verbal thought have disappeared (Vygotsky
presented as arguments for “inevitability.”
1962, 3, Vygotsky’sitalics).
 256 Chapter Twelve
Complementarity as Metaphor 257
foundation of a deterministic description has disap
peared” (1955a, 90 There is nothing essentially quantum in this argument; the “whole-
myitalics).
— ness” seems to apply to anystatistical theory, classical or quantum.
_ Yet quantum wholeness does not necessitate either
complementar- Pauli’s wholeness, as well as Bohr’s, is not a crisply defined concept. For
ity or indeterminism. Bohm’s causal ontology
of the quantum world both Bohrand Pauliit is a tool for the legitimation of the Copenhagen
is an eloquent counterexample to Bohr’s catego
rical assertions. Even credo. ForPauli, the notion of wholeness possessed considerable appeal
though both Bohr and Bohm emphasized the contex
tual nature of as a mystical idea. Pauli attempted to connect the physical realm and
measurement, their notions and uses of wholeriessa
re radically dif- the unconscious, breaking the limits of rational knowledge. He specu-
ferent. Bohr, turning away from mathematics, talked
about the “essen- lated that the humanistic and irrational concept of anima mundi (world
tial unsurveyability” of the measurementinteraction. Bohm
considered soul) should replace the spiritually empty and conceptually bankrupt
measurementto be a special case of the quantum proce
ss and subjected ° mechanical conception of the world as a great machine (Laurikainen
it to detailed mathematical analysis. Bohr, using theno
tion of whole- 1988, 42-43). Pauli welcomed thestatistical nature of quantum me-
ness, deniedthe possibility of a space-time description of the
two-slit chanics because it opened the way for “irrational causes’”—those con-
experiment or a space-time description of energe
tic transitions in the nections between things that cannot be subjected to rational analysis.
atom. Bohm provided causal, realistic space-time descriptions
of both It also opened a possibility that “supernatural will,” in the sense of
(Bohm and Hiley 1993; Cushing 1996). Bohr based his conclu
sions on Schcpenhauer, permeates the world and determinesits actuality (Lauri-
the breakdownofthe classical idea of motion and on the allege
d im- kainen 1988, 55). While Pauli had someinitial reservations about the
possibility of coming up with entirely newideas. Bohm const
rued a ‘Jungian idea of synchronicity (coincidences that are experienced as.
new conceptualtool for describing nonclassical space-time behav
ior— mystical or supernatural and that are not described by the ordinary .
the “quantum potential.” Bohm’s wholeness is a mathematicaltool
for concept of causality), he nevertheless was. driven to broaden “the old,
the exploration of the quantum world. Bohr’s wholenessis a weapo
n of narroweridea of causality (determinism)” to the more general form of »
prohibition. , - “connections” in nature. In this effort, he was especially keen about
The connection between wholeness and indeterminism is also
promi- “correspondence,” or “meaningful coincidence,” between the comple-
nentin Pauli’s writings. As with Bohr, so too with Pauliit is often
not mentarity or wholeness of quantum physics and the idea of the uncon-
clear what follows from what, whatis being assumed and
what de- scious (Laurikainen 1988, 204). He advanced imaginative, metaphorical
duced. In his later years, Pauli proposed an argument,
according to ideas of wholeness and acausality to a mystical abyss that Bohrfelt no
which inseparability between the observer and the observed
phenom- inclination to approach.!”
ena is adirect outcomeofthestatistical nature of quantum
mechanics: The post-Bell understanding of quantum inseparability, or whole-
“If two observers do the same thing even physically,it is, indeed
, really ness, relies on a mathematical analysis of previously interacting, but
no longer the same: only the statistical averages remain in genera
l the now separated atomic systems. It was the “conservatives,” Einstein
same. The physically uniqueindividual is no longe
r separable from and Schrédinger, who provided the first adequate analysis of quantum
the observer” (quoted in Faye 1991, 196). Like Bohr’s metap
horical inseparability by analyzing mathematically the features of the corre-
allusions to wholeness, Pauli’s argumentis intuitively
appealing yet sponding wave functions in such situations.'* Bohr’s translationof the
conceptually fragile. The notion of the “same experiment
al arrange- EPR mathematical analysis into his own terms obscuredrather than re-
ment” in which different experimental results are
obtained, or of an vealed quantum wholeness (chapter 7). Quantum inseparability was
“observer” that “does the same thing” as another
observer, is never buried in a “tranquilizing” metaphorical web spun through positiv-
adequately defined. This notion presupposes, rather than
argues, the ist maneuvers and the repeated enunciation of “inevitable” comple-
impossibility of “hidden variables” and the inevitabilit
y of indetermin- mentarity and acausality. Only after the mathematical analysis of Bell
ism. Similarly, Bohr’s arguments, such as the
following one, simply beg
the question: “The very fact that repetition of the same experi
ment. . 17. For Born “wholeness” had a well-defined meaning: in quantum physicsthere is
in general yields different recordings pertaining to the object
, immedi- “wholeness” because to predict a situation one needs to knowthe wave function every-
ately[!] implies that.a comprehensive account of experiencein where, including the boundary, where conditions are determined by the specifics of the
this field
mustbe expressed bystatistical laws” (1958c,4). experimental situation (1950, 101).
Pauli’s argumentfrom acausality to wholenessis also uncon 18. Einstein, Podolsky, and Rosen (1935); Schrédinger (1935). See Fine (1986) for a de-
vincing. scription of Schrédinger’s crucial role in the elaboration of the idea of “entanglement.”
 258 Chapter Twelve
Complementarity as Metaphor 259
(1966) did progress again become possible. Some scholar
s attempt to how the explanandum gets built into a mathematical model (Hughes
construe Bohras “anticipating” Bell. But “anticipation” is a biased
his- 1989). This is exactly what Bohr refused, or was unable,to do.
toriographical notion, with Whiggish assumptions built into its
very
meaning. :
Somephilosophers argue that because Bohr did not accept nonloca * Bohr: Mathematics and Common Language
l-
ity, he must have embraced inseparability (entanglement): Bohr
must WhenJohn Slater came to work with the great Bohr in Copenhagen, he
have denied that “separate systems possess separate real
states” (Folse discovered, to his amazement, that Bohr had no use for mathematics:
1989, 264). Yet nothing in Bohr’s writings warrants
such a conclusion.”
Bohr's frequent assertion that in atomic physics “wholeness “I had supposed, when I went to Copenhagen,that although Bohr’s pa-
is going pers looked like hand-waving, they were just covering up all the mathe-
beyond the ancient idea of the limited separability of
matter” is too matics and careful thought that had gone on underneath. The thing I
vague to be interpreted in terms of post-Bell inseparability
. Bohr used convinced myself of after a month was that there was nothing under-
this expressionin different conceptual circumstances, from
discussing neath” (interview with Slater, AHOP). What was “underneath”. was —
the initial idea of quantization, through propoundingthe ideaof
dis- Bohr's intuitive analogy-based approach to science. It enabled Bohr to
turbance, to declaring priori the unanalyzability of the quant
um in- guide his mathematically competent students to elaborate rigorously
teraction. Bohr’s intuitions are congenial neither to nonlocality
nor to Bohr’s analogies by use of the correspondence principle, leading to the
nonseparability. Thus, in the case of the double-slit experiment,
Bohr formulation of a new quantum theory. After the erection andelabora- :
claimed thatif we were able to give a more detailed deterministic
de- tion of the new quantum mechanies, the metaphorical approach en-
scription, we would “really be lost,” for “how could the pheno
mena abled Bohr “to familiarize” himself with the “epistemological lesson”
depend on whetherthishole was open or closed?” (1930, 144).
That Bohr did notanticipate Bell is witnessed by the unease and ea- of the theory without mastering its mathematical formalism. Thereis a.
gerness with which the implications of the Bell theorem were (and mythology about Bohr’s “superhuman intuition”—hisdirect commu-
still nion with the “deep truths” of nature achieved without the mediation
are) explored. Somescholars see the explanation of quantum interco
n- of mathematics.”!
nectedness as “the grand task” of our “golden age of metaphysics”
(Shimony 1989; Teller 1989). Many understandthatit cannot be done Niels Bohr’s brother Harald, a great mathematician, once said that
in ’ ‘Niels ‘could get along without mathematics because of his great intui-
Bohr’s way, by keepingtheclassical “words and pictures without keep-
ing the meaning of the words and of the pictures” (interview with tion. Heisenberg declared that “mathematical clarity in itself had no
Hei- virtue for Bohr.” “Physical understanding,” according to Bohr, “should
senberg, February 1963, AHQP). Those who, unlike Fine (1989)
and van precede mathematical formulation” (quoted in Blaedel 1988, 111). Ac-
Fraassen (1989), are not content to accept the foundational
status of cording to Pauli, Bohr himself admitted that his “interest in physics was
quantum correlations” turn to the mathematical models of the quan-
tum formalism lookingforsignificant interrelations. For Bohr, “any ex- not so ‘much that of a mathematician as that of a craftsman and of a
planation is an analogy.” Mathematicians and mathematically incline philosopher” (1964, 1052). Many scholars who analyze Bohr’s work
d notice his peculiar attitude toward mathematics. And, strangely, they
philosophersthink otherwise. They assumethat quantum theory
tran- present Bohr’s choice to dispense with the power of mathematics as a
scendsall classical analogies and look for answers in the
characteristics legitimate stand, if not an advantageous one (Kaiser 1994).
of the mathematical structureitself. Such structural explan
ations are I argue otherwise. There ‘is ample evidence that Bohr’s mastery of
aimed at analyzing invariant connectionsin different models
, display- mathematics was very limited. Lacking advanced mathematical skills,
ing how the elements of a model fit together. Discardingal
l classical Bohr could not build a new quantum ontology but instead hadto use
analogies, one gets an understanding from within, by demons
trating “common language” and simple analogies. This personaltrait, if not
19. Bohr’s assertions of the wholeness of experimental
arrangements are, in a fact
weakness, was canonized into the universal doctrine of the indispen-
statements about the contextual nature of measurement.
For Bohr (though not fee Bohm), sability of classical concepts and the impossibility of a quantum on-
contextualityis different from entanglement because of the tology. Bohr claimed that the quantum world, no matter how muchit
Bohrian “ansurveyability” f
the measurement interaction.
20. One need not ask whythere is quantum holism ven transcends the classical world, could be apprehended only by simple
any more than in Newton’s time
oneneeded to ask whythere exists a gravitational force.
21. Some philosophersof science accept this mythology (Honner 1987).
 260 Chapter Twelve
Complementarity as Metaphor 261
analogies with familiar domains, most of them
nonmathematical. Theo- provided the intuition, Rosenfeld the mathematical calculations: “Bohr
ries of classical physics and their mathemat
ical structures seem to have wanted somebody who knew mathematics and was willing to go
been, in a rather limited way, the only
tools at Bohr’s disposal. Bohr throughit; Rosenfeld sort of sacrificed himself” (interview with Felix
made no contributions to relativity
theory (relativit y played a minor
role in Bohr’s development; Pais 1991, 14), Bloch, 15 May 1964, AHQP). The division of labor was also clear during
nor is there any evidence, | the
will shortly argue, that Bohr really mastered Bohr-Rosenfeld lecturing tour to Russia—-Bohr lectured on “mat.
the formalism of quantum
mechanics. :
ters of principle,” Rosenfeld on the more “technical” aspects (draft of
pS - ee
During the 1930s, the physicists at Bohr’ lectures, AHQP). .
s institute in Copenhagen Similarly revealing is Bohr’s reaction to von Neumann’s remarks on
joked that Bohr used only three mathematical
symbols: >> (much the issues of acausality in quantum mechanics and hiddenvariablesat
greater than), << (much less than), and ~ (about
the same as: Weiz- a conference in 1938. While Bohr (1939) expressed his admirationfor
sacker 1985, 186). Bohr’s brother Harald,
despite being younger than
Niels, had helped Niels in school (Blaedel von Neumann’s mathematical skills, he considered all von Neumann’s
1988, 17). Apparently, Har- mathematical virtuosity superfluous. Bohr insisted. that his own ele-
ald also helped Niels with his epoch-making
,vet mathematicallyquite
simple 1913 paper. According to Oskar Klein mentary analysis of simple experimental cases demonstrated the same
, Bohr's later assistant, all essential point as von Neumann’s elaborate mathematics.
the mathematicsof Bohr’s great 1918 paper on atomic theory was
by Bohr’s assistantat the time, Kramers: “Boh done According to James Franck, Bohr was really “something different”
r had no idea howto do * from ‘other quantum physicists: “Namely, Bohr was an amateur, and
the mathematical part, éven though the physical ideas, so to
had.” In fact, Bohr was somewhat embarras say, he if [a] man like Wigner or so . . . came with too: much involved mathe-
sed when Kramers agrived matics, he even left, and said ‘I can’t understandit’” (interview with
to work with him. According to Klein, Niels asked Harald:
shouldI do with this learned mathematician “What Franck, AHQP).
whois proposing to work Harald provided mathematical help to Niels throughout their lives.
with me?” (interview with Kein, 20 February
1963, AHPQ
;see Dresden As early as 1912 Niels wrote to Harald: “I have thought of you often
1987, on Kramers’s critical role in the elaboratio
n of Bohr’s early atomic these days; for, I had to use some mathematics, and thoughtof asking
theory).”
. for your advice” (Niels Bohr to Harald Bohr, 12 June 1912, BCW, 1:555).
‘Pauli, in a 1927letter, asked Klein to expla
in the conte
nts of the let- And,later, one could hear Bohr saying: “I can’t work out the math, but
ter to Bohr—for he, Klein, was the “only one
in Copenhagen capable here comes Harald” (Blaedel 1988, 18). It was probably Niels Bohr’s
of following mathematical calculations” (Paul
i to Klein, 18 February remoteness from mathematical labor that allowed him to make such
_ 1929, PC). The only interpretive paper of Bohr’
s thathascalculations is
the joint paper with Rosenfeld on the measurab bizarre statementsas“after all, general relativityis just common sense”
ility of electromagne- (1937d, 333). In numerous notes prepared by Bohr for lectures on
tic field quantities (Bohr and Rosenfeld 1933)
. For this paper, Bohr quantum philosophy, nowhere do we find any quantum mechanical
calculations—at most only formulas expressing the uncertainty rela-
22. This state ofaffairs inevitably led to quest
ions about therationale for collaboration tions. What we frequently encounter instead are rough mechanical
and tension aboutthe distribution of credit.
Thus Bohr wrote to Kramers: “One may ex- drawingsof springs, shutters, and bolts for discussions of thought ex-
pect to obtain a large number of significant
results in the nearest future; henceit is dkee.
lutely necessary that both of us are perfec
tly clear about the form of our collaboratio periments (notes for talks at Caltech, at Institut d’Henri Poincaré, and
and that this be arrangedin a way that is reaso n, in Japan, 1937, AHQP). Discussing simple thought experiments, Bohr
nable andjustifiable for both of us i
talking with youlast night about calculations informed his audience that he “tried not to enter into any details of
, I got the feeling that perhaps you ae -
think that the manner of continuing our colla
boration . . . is the wisest for you, butthat the mathematical methods” (1937d, 267). Yet all of the available evi-
you mightpreferto try, more independently,
to work out some problem.” Kaanriers felt dence, published as well as unpublished, showsthat Bohr never devel-
uneasy, yet reassured Bohrabout the impor
tance of Bohr’s contribution: “If Tonly knew
well and clearly what the matter is. The whole oped any mathematical treatment of these thought experiments. Nor
thing appears to me as some vague half
ethical and half practical question . . . just did Bohr systematically elaborate his brief (too brief) analogies between
as little as I can refrain from working little
independently,just aslittle can I refrain from
the matter so very much by the philo
later asking your advice, for you can furthe
r
quantum physics and other domains.
sophythat you can put intoit” (Bohr
15 November 1917, Hellerup, Niels Bohr Archi to Kramers According to Heisenberg, a consistent, rigorous interpretation of the
ve; Kramers to Bohr, 11 November 1917, quantum formalismis not possible without recourseto its mathematical
Niels Bohr Archive, quoted in Kojevnikoy
1997).
‘ aspects: “One maysaythat the concept of complementarity, introduced
 262 Chapter Twelve
Complementarity as Metaphor 263

by Bohr. . . has encouraged physicists to use an ambiguousrather than

about everything, that finally he became always ili and stayed home
unambiguous language,to useclassical concepts in a somewha
t vague when the day came to write essays” (interview with Franck, AHQP).
manner. ... When this vague and unsystematic use of language
leads Heisenberg singled out Bohr’s difficulty with writing as one of the rea-
into difficulties, the physicist has to withdraw into the mathematical
sons for Bohr’s early withdrawal from active scientific research. Ac-
scheme”(1958, 154).
cording to Heisenberg, when he came to Copenhagenin 1925, Bohr had
This sort of withdrawal, I have argued, was not open to Bohr, who
already stopped doing physics. Heisenberg found Bohr overwhelmed
never seemed at ease with the esoteric game of symbols that his
with administrative duties: Bohr “had many visitors and then writing °
younger colleagues had mastered effortlessly. Analyzing the use of
letters. . . . Since it took him so much time to write even a single quite
analogy by alchemists before the “mathematization of the world pic-
trivial letter, it’s obvious he could notfind time for many other things”
ture,” GentnerandJeziorski, following Vickers (1984) wrote that alche-
(interview with Heisenberg, AHQP). SB ag
mists were “ownedbytheir analogies, rather than owning them”(1993 Part of the Bohrian myth is that he thought very clearly and only
448). Can we not makea similar assessment of Bohr? ‘ expressed himself obscurely. Bohr’s struggle with languageis often pre-
sented as an attemptto encapsulate his worked out, but not yet spelled
Metaphorical Appeal and Conclusion
out, ideas. The ideas themselves, so the myth goes, are sharp. crisp, un-
ambiguous; only their “expression” is vague. Bohr himself strength-
In later years Bohr based his notion of scientific objectivity on nonam- ened this impression: “I must have expressed myselfmost confusingly”
biguity of communication. Yet the metaphorical nature of complemen- (1938b,31). This separation between Bohr’s thought and his language
tarity implies its essential ambiguity.It is not clear how Bohr’s emphasis seemsespecially out of place in view of his emphasis on the holism of
on nonambiguity of communication is compatible with his insistence mental phenomena. Thedistinction between Bohr’s linguistic formula-
that the “definition of a concept stands in complementary relationship
tions and unexpressed thoughts is peculiar when advocated by Bohr’s
to its: practical implication.” Evenless clear is how such nonambiguity followers, who subscribe to his dictate that no phenomenonis a phe-
ITT OT

is possible when “improvisation . .. points to a feature essential to all

nomenonunlessit is an observed phenomenon. From this perspective,
communication” (Bohr 1955b, 79).
what access to Bohr’s thought could one have except through Bohr’s
Duality betweenthe rhetoric of nonambiguity and the essential am- talking arid: writing?
biguity of improvised, associative thinking produces anothertension in From the early days of his scientific career, Bohr dictated his papers
Bohr’s writings. It is the frequent vacillation between the two extremes
and scientific correspondence to numerous assistants. Their role was
spreereinemere

of presenting his complementarity philosophy as exceedingly compli- to help Bohr with the formulation of his ideas, to provide mathemati-
cated or as transparently obvious. Thus Bohr announcesthat “the dif- cal help, and to keep the discussion within the confines of relevancy.*
ficulty is, of course, very, very deep by nature” (1930, 136). Explainin
g Bohr’s assistants did not, of course, simply record Bohr’s fully backed
complementarity and claiming that “phenomena are essentially the re- ideas—ratherthe ideas were created in.a dialogical, yet unequalinter-
sult of the interaction with the measuring instruments,” Bohr warned change between Bohrandhis helpers. Bohr’s assistants often had no
his listeners that “this is very important. . . very complicated” (1937d
aes

choice but to help Bohr to formulate ideas they themselves disagreed

262). Yet a few lines later he called it “very simple’: “It is something
with. Reminiscences by Kalckar give the flavor of this process, as well ©
very simple—thesituation can be explained and can be
detailed by as of Bohr’s personality: “Sometimes, during work on a particularly dif-
looking a little closer into some of the very simplest experimen
ts” ficult point in a treatise, where one had not succeeded in reaching any
(1937d, 263).No wonder Bohr shuttles between the opinionthat physi-
unanimityinthe final formulation, and when,therefore, in orderto get
cists have readily accepted complementarity——“it is something very
on, one hadacceptedhis suggestion without really approving it in ones
simple which at the momentby physicists [sic] appear quite clear”—to
the opinion (just two sentenceslater!) that “all physicists will have to heart, he might stop walking around the table. He would look pene-
work very hard to get accustomedto [it]” (1932, 201).
Bohr was notorious not only for the incomprehensibility of his talks 23. Rosenfeld recalled helping to draft Bohr’s replies to Pauli’s letters: “I would stand
butfor his extreme difficulties with writing. As Franck recalled: “Bohr byas usual, in myrole of helper, putting a hand to the more mathematical considerations
had so great trouble to write little essays one has to write in school and generally signifying by approbation or dissent whether the statements diffidently
proposed by Bohr seemed or not to meet the point under debate” (1967, 119).
 264 Chapter Twelve
Complementarity as Metaphor 265
trating at one and say: ‘Now, you mustn’‘t
be unhappy about it. Don’t seems to methat Bohr’s analogies, insofar as they pertain to something
think I’ve gone quite crazy, but...’ And then he
would go over the psychological, are an arbitrary construction without any deepersignifi-
disputed argument from every side . . . until hefel
t that his young pu- cance” (Pauli to MarlusFierz, translation in Laurikainen 1988, 200, my
pil was again with him” (1967, 230). In fact, in each
paper signed by italics).
Bohronly, an informed reader can perce
ive theintellectual inclinations Metaphors often underpin both commonand specialized scientific
of the assistant who helped in its compo
sition. The two versions of the language (Lakoff 1993). Terms that seem to be purely abstract have an
Como lecture, written only weeks apart, form
an eloquent example. The underlying metaphorical meaning (Turner 1987). Such supposedly neu-
manuscript of the lecture (Bohr 1927c), with its
emphasis on the wave tral terms as “causality,” “disturbance,” and “interference” may trigger
nature of matter, bears the imprint of Darwi
n’s predilection and pref- a leap to desired, unfounded conclusions. When Donnan wrote to Bohr
erenc e for Schrédinger’s wave mechanics. TheNature
1928), written with Pauli’s help, has a much version (Bohr that his complementarity reasoning, following from the disturbance
stronger operational em- idea, was absolutely compelling, he was seduced bythe vivid attrac-
phasis. : : oe tions of metaphorical imagery (Donnan to Bohr, 1935, AHQP).
The obscurity and vagueness of Bohr’s writing reflec
ts the associative Yet the wide appeal of complementarity is based on its existential
and metaphoricalnature of his thought. Numerous
anecdotestell about message—anaffirmation of the “irreducibility” of life, an invitation to
Bohr's talking before thinking, or Starting a sente
nce without knowing sustain and to accept contradictions,» a quest for unity that transcends
howto complete it. According to witnesses, Bohr would
sometimessay: the fragmented, painful human existence. The appeal of complemen-
“I do not know howto finish this sentence.” * Bohr’
s brother Harald tarity is in its proclaimed goal of “harmonizing ‘all situations” (Bohr
his loyal supporter, explained the Opaqueness of expression
: “Niels 1958e, 699)— from contradictions in physics (wavesversusparticles) to
talks about things he means to explain later” (Pais
1991, 45). But most contradictions in spirituallife (the pragmatic versus the mystical; Bohr
things are never explained later—Bohr offers confi
dent promises that 1960c, 14-15). When Weizsacker found himself enchanted by comple-
are not, and cannotbe, fulfilled. Among such promi
ses belongs Bohr’s mentarity, it was because it promised to bridge his scientific and per-
claim that his complementarity philosophyprovides
a resolution ofthe sonallives (1985, 184-85). Bohr’s discussion of existential matters reso-
mind-body problem. Reading Bohr, one is naturally
led to assumethat nated strongly in the tender souls of his young disciples: “Best ofall
somewherein his writings or unpublished notes
, one canfind the fully was whenthe conversation turned to the so-called eternal questions.
worked out solutions. Yet the solutions do not exist.
Bohr schol ars Nowhere did Bohr’s influence as an educator have a more profound
therefore charitably and ingeniously reconstruc
t from Bohr’s short re- significance. . .. Nowhere was the overpoweringintellectual and emo-
marks what he might have had in mind. Each scholar
inevitably brings tional impactof his personality more irresistible” (Kalckar 1967, 235).
in his or her own, not Bohr’s, set of assoc
iations. Perhaps becau se of WhenBorn wasdistressed by the Cold War and devastated by the
Bohr’s open-ended analogies, Pais coined the term “com
plementarism” use of the atomic bomb on Japan, he turned to the philosophy of com-
to distinguish them from what he supposed to be Bohr’
s logically im- plementarity because of its conciliatory message (chapter 8). For Bohr
~ _peccable and unambiguous complementarity
in physi cs (1991, 438-47).
The analogies of complementarity are Partial, himself, his numerous analogies had a profound emotional signifi-
incomplete, vague, and cance. He was constantly struck (“striking analogies” —Bohr 1955a, 93)
contradictory. This is not surprising, for, accordingt
o Bohr, the main bycontrasts around him, and he commented on them sometimes ina
concl usions of his complementarity philosophy follo
w from the indi- sage and sometimes in a humorous way.”* Bohr hadanintenselifelong
visibility of the quantum ofaction, yet there is no rigor
ous analogue of attachment to complementarity, as Kepler had to his cosmic geometri-
this indivisibility in nonquantum domains. Conse
quently, no mean-
ingful structural network of analogies can be
established betw
een say, 25. According to Kierkegaard, the ability to sustain conflicts rather than trying to re-
quantum physics and psychology. Pauli, the coaut
hor of complemen- solve them belongs to a higher“stage”oflife.
tarity, must have understood the limitations
of Bohr’s analogies: “It 26. “The mutually exclusive relationship between different psychological phenom-
ena, denoted by such wordsas ‘thoughts’ and ‘feelings,’ or ‘reason’ and ‘instinct, remind
24. To whichDirac apparentlyreplied: “I us in fact most strikingly of the complementarity relationship between atomic phenom-
was told never to start a sentence which
does not knowhowto finish” (quoted in French one ena described by kinematical and dynamical concepts respectively” (Bohr 1939, 403, my
and Kennedy1985, 249).
italics).
 266 Chapter Twelve
Complementarity as Metaphor 267
cal construction of the plan of the
Creator. As Nietzsche lived his
philosophy, so did Bohr. After 1927 Bohr systematic, rigorouselaboration ofthe idea of complementarity.?” With
would talk for hours about the
significance ofthe “great connections” he the passage of time, Bohr’s presentations of complementarity became
had discovered.
Part of the appealof complementarity more, rather thanless, ambiguous. This ambiguity allowed verydiffer-
has to do with ideas so deep]
entrenched in western European culture that one wrongly cousd ent gudiences to experience kinship with the philosophy of comple-
them unquestionable, obvious, universal. ilers mentarity — from Jordanin his Nazi days (Beyler 1996; Heilbron 1987;
Complementarity between
“thought”and “feeling” is, I have argue Wise 1993) to today’s New Age advocates,left intellectuals (Aronowitz
d, such an example. Another
especially “liberating” (Pais 1991, 1988), and feminists (Wilshire 1989). A rigorous clarification and anal-
447) example is that between love
and “justice.” As the “mutual exclu ’ ysis would undermine Bohr's metaphorical framework. Davidson’s
siveness” of emotion and reason,
the complementarity of love and justi so words (though relating to another matter) aptly characterize Bohr’s
ce draws its appeal from con-
trasting the just and therational (male) with metaphors: “It is hard to improveintelligibility while retaining the ex-
the emotional and thein- citement” (1984, 183).
tellectually’ inferior
(female). The eloquent “argument” - ;
complementarity is that of Herbert Spen behind the In oneofhis later talks Bohr referred to complementarity by telling
cer (1873): “The love of the
helpless inevitably affects her [woman's] an amazing “little joke”: “Churchill said, about theBattle of Britain, that
thoughts and sentiments, and
this being joined in her withthe less deve never in history have so fewbeen able to do so much for so many. Now
loped sentiment of abstract
justice” (quoted in Easlea 1981). The great wecan certainly turnthis around in physics and say that hardly, in the
est scientists, Bohr and Ein-
stein included, accepted uncritically manyof the history of science, have so manyphysicists . ..worked so many years .
stereotypes of their
culture. Bohr felt that a “specially drastic example”is and so hard on so small and simple [a] point” (1958e, 705). Wasit
the one in which
“the role of men and womenare reversed” (1938b, 30). the epitome of Bohr’s numerousvacillations between complementar-
The power of complementarity as a weapon of legiti ity as difficult and profound and complementarity as simple and self-
mation flows evident? Or might Bohrhavebeenteasing thecollective of future schol-
from its emotive, existential, and
cultural appeal. Complementarity
was indeed more warmly embracedin Europ ars who have invested so mucheffort and ingenuity in presenting his
e than on the more prag-
matic American soil (Cartwright 1990, 417-24). Bohr used ambiguousset of associations as a rigorous philosophical framework?
his brief

‘ full Bs
analogies with other fields less to explore and
knowledge than to prep
elabo rate the unity of © " +27: Bohr was awareof the frustration his audiences often experienced: ‘You might
are audiences to accept compleme
physics: “I just wantto explain whyI ntarity in mostof you think that I am usinga lot of wordsandthatjust plain words and howwill I
think it has importance to empha- define these words? But now ofcourse thatis right, that is just the whole difficulty of
size the analogy. First ofall it is inter
esting when one as [sic] regards speaking about any field without thatit were verydaily experience’ (1937d, 354).
the situation in atomic physics. . . .
it is, of course at any rate, very
alarming that we should be in this situa
tion . . . butit is interesting to
remember that there are otherfields of
knowledge wheresimilar situa-
pail must be looked into and have been
looked into” (1937d[lecture 6]
52-53).
Bohr, despite his modest (and theto
rically pointed)
declarations that
he “has been forced to enter on
such problems” only for his “ow
ness of mind” (1937d [lecture n clear-
1], 263-64), made deliberate
shape his presentations in the most efforts to
persuasive way. Thus he informed
his former assistant Kramers abouthistal
k at the CopenhagenSociet
for Philosophy and Psychology:“I lear
ned a great deal from the suas
ing discussion. In particular, I know
better which points nonphysicist
resent, and I also eeve that for s
this very reasonI found onthis
Se re
sion bett er words than p previouslyto occa -
y answerthe obje ject ions
ctio ns ” (quoted in
;
When Bohr adjusted his talks
to hig audiences, he did not
present a
 CHAPTER 13
AS
Hero Worship, Construction of Paradigms,
and Opposition

Even to the big shots, Bohr was the great God.

Richard Feynman 1980, 129

Scientists are inclined to take their own outlookfor the natural way of looking
at things, while the outlookof others, inasmuch as they differ from theirs,
are adulterated by preconceived and unwarrantedphilosophicaltenets,
which unprejudiced science must avoid.
Erwin Schrédinger 1955, 14

Introduction

In our times, the dethroning of heroes is commonplace. Weare reas-

sured to learn that our cultural and scientific heroes—the overpower-
ing Newton and incomparable Mozart—are in fact flesh and blood, en-
dowedwiththefrailties, ambitions, and vices of simple mortals. In the
history of science we also have come to recognize hero worship as a
major distortion of the past.
Yet hero worshipis not merely a historiographicalsin. It is a powerful
toolfor legitimating orthodoxy. Hero worship does not merely deify and
=
ad mystify the hero. Hero worship erects psychological barriers against
Fes
= criticism and delegitimates opposition. Hero worship effectively con-
ceals the dialogical network in the emergence of novelty by centralizing
insights, dispersed among manyparticipants, into the handsof a few
greatscientists. In this way, the “lesser”figures are painlessly eliminated
from history.If, in the evolutionary narrative, the scientific community
becomes superfluous, in the revolutionary account, the scientific collec-
tive is no more thana dutiful army of puzzle solvers. The communica-
tive, interactional nature of scientific creativity is as alien to the revo-
lutionaryas to the evolutionary narrative (see chapter 15).
In this chapterI describe what can be called the hero worship of Bohr,
 270 Chapter Thirteen
Hero Worship, Construction of Paradigms, and Opposition 271

and I indicate howBohr’s authority promoted uncritical

acceptanceof arguments of a lecturer pertaining to isomers were “all wrong” (1991,
the Copenhagen philosophy.I also describe the strateg
ies by which the 7). Reminiscences, irrelevant to the scientific problems at hand, Pais
orthodox portrayedtheir insights as revolutionary by discrediting the
presents as “warming up exercises,” while Bohr’s inability to deliver a
opposition and constructing a simplistic notion of past science.
coherent public lecture Pais interprets as an “unrelenting struggle for
Bohr and Hero Worship
truth” (1991, 11).
To a certain extent, Bohr wasa tragic figure. He laid the foundations
One cannot overestimate the impact of the authority figure in the of the quantum theory of an atom andinspired.and supervised the erec-
tion of the new quantum theory. Yet-he was unable, withhis heavyad-
evaluation and acceptance ofideas. Bohr’s unprecedented authority not
ministrative duties and limited mathematical knowledge, to participate
only promoted the widespread, uncritical acceptance of the Copenh
a- actively in further developments whenthefield became too mathemati-
gen philosophy but obtained a favorable reception for
his dubious and cal (chapter 12). While Bohr presented himselfas a dilettante who had
poorly developed ideas outside of his area of competence
(Heilbron to approach “every new question from startingpointof total igno-
1987). Bohr’s deliberations on complementarity in biology,
psychology, rance,” Pais graciously remarks that “Bohr's strength lay in his formi-
and anthropologyare good examples.!
dable intuition and insight rather than in erudition”(1991, 7).
Bohr’s authority was based onhis outstandingscientific achiev
ements Bohr’s “formidable intuition” and “subtle reasoning” were often used
in the past, his formidable institutional power, and his unique personal
by. the orthodox to certify the Copenhageninterpretation as final and
charisma. Bohr could provide simultaneously intellectual stimula
tion to disarm those who soughtanalternative. The legend that Bohr had
and help in advancing careers, spiritual fulfillment and down-to-earth
some sort of access to nature’s secrets, qualitatively different from that
fun, material benefits and psychological counsel.2 Bohr became a father
of other mortals, directly discouraged critical dialogue. This legend is
figure whose unique status many young scientists were eager, to honor
supported by another, peculiar claim—unlike other theoretical physi-
and whose authority not many daredto challenge.
cists, Bohr did not need to calculate in orderto obtain “the truth.” Blae-
Many physicists, among them Pauli and Jordan, used the words
del’s is a typical statement: “Perhapshis intuition allowed him to grasp
“father figure” whenreferring to Bohr. Pais (1991) calls Bohr a “father
things when others neededcalculations” (1988, 11). Bohr’s personal
figure extraordinary.”Jordan referred to Bohr as “the fatherof the great
limitation is thus uncritically transformedinto a strength. At the same
world-wide family of quantum physicists” (1972, 214). Bohr was afa-
time, strong legitimiation is given to Bohr’s metaphorical presentation
ther figure for “physicists belonging to several generations”
(Pais 1991, - of complementarity by singling out his associations as privileged (chap-
3). Bohr’s assistant Kalckarcalled Bohr a “fatherly friend” (1967, 227).
ter 12). Needless to say, none of the cases in which Bohr turned out
Scientific heroes are often described as being endowed not only with
to be wrong—the Bohr-Kramers-Slater theory, the existence of spin
supernatural reasoning faculties but also with superhuman personal
and of the neutrino, complementarity between “vitalism” and ““mecha-
qualities. Any eventor trait, no matter howirrelevant, or even discred
- nism” *—are mentioned when Bohr’s intuition is being worshipped or
iting, is construed as reinforcementof the hero’s uniquestature. Often
such a construction is accompanied by
moo used as a weaponagainstdissent.
genuine affection, as in Pais’s
It became almost obligatory,when writing about Bohr, to refer to the
(1991) depiction of Bohr. Pais presents Bohr’s ignorance as formidable
intuition —Bohr did not know what an isomer was “subtlety” of his thinking. Followers and opponentsalike characterized
but knew that the Bohr'’s thoughtas subtle. Yet, peculiarly,it is rarely a specific argument
1. As Landé commented: “It is a matter of social psychol
that is singled outfor this description. Rather, the word “subtle” is used
ogy how people perceive
the whenthe author encountersa difficulty in understanding Bohr, failing
Ao uttered by great men outsideof their field of expertise” (intervi
ew with Landé to penetrate the structure of the argument, to achieve a clear and coher-
P).
‘
2. All this was offered at Bohr’s institute in Copenhagen.
As early as the 1920s, Bohr ent reading of Bohr’s writings. Thus Sklar, characterizing -Bohr’s phi-
could support Heisenberg’s and Pauli’s visits to Copenhagen muchbet losophy as “exciting and subtle,” simultaneously complains that “his
ter than the young
German physicists could expect to be financed in Germany
funding agencies, including the Rockefeller Foundati
. About Bohr’s connection with so-called CopenhagenInterpretation is not easy to summarize neatly”
on, as well as about the institute
as a place of fun (sports, the opposite sex), see Pais (1991). (1992, 172) and_ that “complementarityis a difficult notion to fully pin
Bohr’s expertise in securing
funding is explored in Kojevnikov (1997). Maiy recollections
by Bohr’s colleagues and 3. Bohr avoided this terminologyafter the discoveryin 1953 of the DNA structure by
disciples reveal a unique, powerful charismatic personality. :
Watson and Crick.
 272 Chapter Thirteen
Hero Worship, Constructionof Paradigms, and Opposi
tion 273
down”(173). Hailing Bohr’s Philosop
hy as “extraordinarily ingenious,”
Sklar calls it “not a view of the worl The fact that many physicists were willing to accept Bohr’s
d easy to understand” (1992 175) authority
Even Bohr’s opponentsfelt the Press on fundamental issues of quantum theory is not entirel
ure. Thus Bohm and Hiley, de- y surprising.
scribing “Bohr’s very subtle thou Bohr’s arguments for the consistency of quantum mechan
ght,” mentioned that his ideas ics (his re-
not appearto have been well understo “do spgnse to EPR) and of quantum electrodynamics (Bohr and
od bythe majority of physicists” Rosenfeld
(1993, 15). It seems that the moreone 1933)® were perceived by physicists as a vindication of the power
feels at aloss to extract a clear
and s and
coherent message from Bohr’s phil
osophy, the more often one calls
consistency of their tools, as a green light to go on confidently (Beller
“subtle” and “ingenious.” This it and Fine 1994; Pais 1991). Few bothered to carefully study either the
terminology creates a qualitative
between oneself and thehero, excu ga EPR paperor Bohr’s response toit?
sing inability to understand arid
multaneously preventingcriticis st Morestriking are testimonies of blind acceptance of Bohr’s author-
m of the hero’s authority. No wond
Bohr’s noti on of “deep truth” was usedto disarm er ity even in such matters as the physics of the atomic bomb. According
tors to Bohr’s institute who thought those young visi- to Feynman’s reminiscences about Bohr’svisit to Los Alamos, nobody
they had found contradictions in
his reas oning#* Any doubts physicists had there dared to subject Bohr’s proposalfor improving the bombto criti-
abou
t Copenhagen wisdom
were silenced by an appeal to Bohr's cal scrutiny. Feynman, who was unawareatthe time that he was talking
unreachable depthof discourse
beyond the usual “simplistic” cannons to the great Bohr himself, criticized Bohr’s ideas freely. After the meet-
of argumentation. When the
“simple” truth seemed to contradict the ing Bohrtold his son Aage that “Feynman had been the only person at
orthodoxy, the inaccessible
“deep ”truth prevented further questioning. the meeting who had been willing to say that an idea of his was ‘crazy’”
This is not to deny ingenuity to Bohr’s (Schweber 1994, 403). Accordingto Feynman, Bohr said: “Next time
reasoning
on complementarity,
in which arguments of consistency are when wewantto discuss ideas, weare notgoingto doit with these [big
subtly disguised as those ofin-
evitability. Someof the consistency arguments are very shots] . .. who say everythingis yes, yes Dr. Bohr” (Feynman 1976, 29).
arguments owe their subtlety to skillfu clever; other Hero worship, and the associated suppression of criticism, need not
l ad hoc maneuvering.> One can
easily criticize the lack of rigor in such arguments, always be sober andserious.* Thus Hendrik Casimir, a physicist from
for Bohr and his Bohr’s circle, wrote a comical poem for Bohr’sfiftieth birthday. He de-
followers often transcended the frame
work of nonrelativistic quantum
s OCT = '

mechanics while attempting to demons scribed Bohr’s famoustheory that a defensive, as opposed to aggres-
trateits consistency. In the case
of the Bohr-Einstein argument about the sive, shooter has an advantage because supposedly making a voluntary
time-energy uncertainty rela-
tion (see Bohr 1949, 224-28), Bohr wasfo decision takes more time thanreacting in a purely mechanical way. The
rced to invoke ad hoc a for-
mula from general relativity theory. In the poem concludes with an attemptto give Bohra gun so he can provehis
case ‘of Heise
nberg’s y-ray
thought experiment, Bohr used the formulas theory “experimentally” by defending himself:
of classical wave theory
while simultaneously denying that theory
ea aT pect

's validity by appeal toa pho- So the three of us wentto the center of town
fon concept. No wonderEinstein was
not swayed by such arguments— And there ata gunshop spent many a crown
it is an illegitimate mixture of classi
cal and quantum concepts simila Onpistols and lead, and now Bohrhadto prove
to the
setup he found so unsatisfactory in Planc r
k’ initial derivation of Thatin fact the defendantis quickest to move.
the distribution law. Some physicists Bohraccepted the challenge withoutever a frown;
, such as Heisenberg and Bopp
understoodthatthe frivolous mixingof He drew when wedrew ... and shot each of us down.
classical and
quantum concepts
‘in the analysis of thought experime This tale has a moral, but we knewit before:
nts was inconsistent, Bohr’s ide-
ology of complementarity notwithstand It’s foolish to question the wisdom of Bohr.
ing. Yet Bohr’s arguments were
accepted uncritically by many physicists Casimir’ translation (Casimir 1983, 98-99): original German
, historians, and Philosophers;
these arguments were construed asthe ' (Casimir 1967, 113)
“triumphantvictory” of Bohr,
heroically countering the threatening move
s of the Opposition.
6. See Schweber (1994) for an excellent discussion of Freeman Dyson’s use of the Bohr
4. According to a saying of Bohr’s and Rosenfeld paper.
, there is a “simple truth,” the Oppos
falsity; and the “deep truth,” the ite of which i 7. “WhenI asked George Uhlenbeck, who wasan active physicist in his mid thirties
Opposite of which is also a truth (1949, 240)
5. Bohr’s uses of correspondence argum ° when EPR appeared, whathe recalled about physicists’ reactions, he replied that no one
ents in his early creative years display
standing ingenuity and skillfulness. Mycri an out- he knewpaid anyattention”(Pais 1991, 430).
ticism here refers to the philosophicalle
mation of the Copenhagenideology. giti 8. I discuss the humorouscritique of complementarity in my “Jocular Commemora-
P =
tions: The CopenhagenSpirit” (Beller 1999).
 274 Chapter Thirteen
Hero Worship, Construction of Paradigms, and Opposi
tion 275
The moral of the story is clear. Nothing positive awaits those who
to be able to tell what he meant and why was he right. | tortured myself
challenge Bohr’s authority. But is not such an interpretation inconclu-
on endless solitary walks” (Weizsacker 1985, 185, myitalics). The ques-
sive and tendentious, misrepresenting the joyful and humorous atmo-
tion was not: Was Bohr right? or To what an extent was Bohr right? or
sphere in Copenhagen?Clearly not, for let us quote Casimir’s own s0-
ber reading of his poem: “The moralof the story that one should On what issues was Bohrright? but, quite incredibly, What must one
not assume and in what way must one argue in order to render Bohr right?
doubt the wisdom ofBohrapplies to more important things than shoot-
ing gunmen in Westerns” (1967, 113; 1983, 98-99).
The exuberant celebrations at Bohr’s institute, such as The Issue of Consistency
a feast held by
the graduate students’ club on the occasion ofthe twenty-fifth anniver-
sary of the institute, are also instructive about the nature of Bohr’s Scholars of Bohr’s thoughtare bewildered by the rumber of contradic.
au-
thority. Students, standing on chairs, beer in hand(with Bohr similarly tions in his writings. As I have argued, the inconsistencies are genuine,
ona chair), sang the hymn “Fathersof Selena,” which hailed the “noble stemming from many conflicting voices, from confusion between con-
Bohr” who “knowsthe wayamidstall false tracks” (Pais 1991, 6). Those sistency and inevitability arguments, from a mixture of metaphorical
false tracks no doubt referred to directions of thought at odds with the and model theoretical argumentation, and from a mixture of classical]
Copenhagenorthodoxy. and quantum concepts. Moreover, unbridgeable gaps exist between
Bohr’s unpublished correspondencediscloses the overwhelming guilt . Bohr’spre-1935 and post-1935 philosophies.
experienced by those physicists who dared to challenge him. Thus Bopp - Encountering these inconsistencies, scholars respond in ingenious
arguedin a letter to Bohr that one cannotobtain a contradiction-free ways and develop different interpretations of Bohr’s thought—any-
interpretation of quantum theory by relying on thought experiments thing short of admitting the contradictions. The unshakable belief in his
that explicitly use the concepts ofclassical particle mechanics—
exactly consistency moreoften than not followsdirectly from explicit hero wor-
the concepts that the new physics gave up. Bopp’s contention strikes, of ship: Bohr’s mind is too “sublime,” “subtle,” “too relentless,” or “too
course, at the very heart of Bohr’s lifelong enterprise. Boppprefaces his scrupulous” to produce inconsistencies (Honner 1987, 12). And even
argument with the following words: “A young Japanese colleague said a historically sensitive philosopher, Don Howard, whoinsists on ap-
once: we must do something, that our parents do not understand, and plying “thecritical tools of.a historian” in order to understand Bohr,
this is very painful” (Bopp to Bohr, 4 February 1962, AHQP).? These defines understanding Bohr as reconstructing “from Bohr’s words a
words, again, reveal the enormous parental authority of Bohr. coherent philosophy of physics” (1994, 201). Howard's imaginative re-
Many of Bohr’s correspondents could not transgress the psychologi- construction, as he does not fail to realize, transcends Bohr’s words
cal barrier of even beginningto criticize Bohr. Thus Jesse Du Mond significantly.?° ,
wrote to Bohr of his numerous unsuccessful attempts to grasp the One mightcriticize such reconstructions of Bohr's thoughtas histo-
meaning of Bohr’s writings: “This is of course not the first time I have riographical sins. And historiographical sins of this sort enhance hero
read and tried to grasp your point of view. ... regarding your
view- . worship not merely by refusing to admitinconsistencies but by ascrib-
point J have never been able to get a clear answer fromothers. ing to the herofigure thelater insights and achievementsof others. Such
I hasten
to say that I do not meanto imply anyshortcomings in your ownwrit- are the arguments for Bohr’s alleged foreknowledgeof Bell’s nonlocali
ty
ten exposition and am very ready to admit that the difficulty is entirely (Folse 1989).
the fault of my own slowness and dullness” (Du Mondto Bohr, 7 March Even the most competent and friendly readers find Bohr’s philosophy
1961, AHQP). z obscure and inconsistent. Physicist and philosopher Abner Shimony
Yet Weizsacker’s reminiscences about his first encounter with admits: “I must confess that after 25 years of attentive—and even rey-
Bohr
constitute perhaps the best evidence for the overpowering, almost erent—reading of Bohr, I have not found a consistent and comprehen-
dis-
abling impact of Bohr’s authority. After the meeting with Bohr, Weiz- sive framework for the interpretation of quantum mechanics” (1985,
sacker asked himself: “What had Bohr meant? What must I understand 109). We might wonder what reason he has to remain “reverent” after

9. “Ein junger japanischer Kollege hat einmalgesagt: wir 10. In fact; Howard uses von Neumann’s analysis based on pure cases and mixtures,
miissen etwas tun, was un- which Bohr himself considered superfluous, and which cannotbe construed as a natural
sere Eltern nicht verstehen, und das schmerzt uns sehr.”
extension of Bohr’s deliberations.
 276 Chapter Thirteen
Hero Worship, Construction of Paradigms, and Opposition 277
repeatedly (for twenty-five years!) finding Bohr’s writings unsatisf
ac- The orthodox exaggerated the difficulties of the opposition stand
tory and whether anybody but a “hero” would receive such charity.
while ignoring their own. As the opponents realized, not without some
Yet perhaps the most extremeexpression of the passion to save Bohr’s
consistency by any means comes from Honner: “If Bohr’s thought bitterness, theircriticism of the Copenhageninterpretation was simply
is “brushed off” with accusations of not “understanding Bohr’ (Landé
not found to provide a consistent framework for the interpretation
of 1965, 123). As in politics, so'in science, the orthodox misrepresented,
quantum mechanics, then perhaps one’s expectationsof ‘interpretation’
trivialized, and caricatured the opposition’s stand.
should be revised” (1987,23). Clearly, the soundness of Bohr’s reason-
Perceiving threats from the technically sophisticated and proficient
ing is not judgedbyobjective standards, rather Bohr’s authority defines theoretical explorations of Einstein, Schrédinger, Landé, and Bohm,the
~ what the legitimate standards of reasoning should be. orthodox translated them into “simple” language. This tactic ensured
As befits a hero, Bohr’s weaknesses -in argumentation are
trans- that few would bother to look into the original arguments, learning
formed into assets. Bohr’s obscurity, Folse explains, followed from
his about the opposition stand through a presentation by the orthodox.
greatness—"a work of geniusresists categorization” (1994, 119): Bohr’s
Such translation not only distorted the original arguments, hidingtheir
difficulty in writing followed, according to Chevalley (1994), from his
strengths, but also tailored them for the upcomingattack. Bohr’s simple
philosophy of the ambiguity of language. According to Honner, the exposition of the EPR argument misrepresented its original structure
opaqueness of Bohr’sphilosophyfollowed from “his concern for preci-
and hid its radical message about either the incompleteness or the
sion of expression”; repetitiousness was a proof ofdepth: “Thearticles nonseparability of quantum. mechanics (chapter 7 and Beller and Fine
are often repetitious, but it only should deepen our convictions that 1994). Diverting the issue from the focus of the opposition’s criticism
they [Bohr’s examples] are not merely analogies” (1987, 169). to a position on which the opponent may be weaker is a strategy of-
Philosophers often apply double standards when judging the ortho-
ten used. Thus, in a controversy between Born and Schrédinger, Born
dox and the opposition. A weaknessin the opposition’s stand might be
(1953b) advanced arguments against Schrédinger’s (1952a) attempts to
a reasonfor total dismissal. A similar deficiency in the case of the win-
eliminate particlehood, when the bulk of Schrédinger’s paper was di-
Coe RPTSGRSRSISAT

ners is downplayed and rationalized. Hooker, one of the most sympa-

rected against quantum jumps. /
thetic and penetrating commentators on Bohr’s thought, after (reluc-
The orthodox did not merely misrepresent the opposition’s stand—
tantly) criticizing Bohr’s doctrine of the indispensability of classical
they trivialized and often deliberately caricatured it. Schrédinger’s at-
concepts, defends him: “Bohr was only driven to adopt indispensabil-
tempts to develop a wave ontology, which originated in his work on
ity [of classical concepts] byhis efforts to understand the conceptual
vibrations in a gas before the new quantum mechanics and wererein-
significance of quantum theory.” Moreover, while there is no evidence
forced by his researches in cosmology andstatistical mechanics in the
yg oiggeeigecoaremetnnannpeand

that Bohr ever considered abandoningthis peculiar, untenable doc-

1930s and 1940s, were the result of penetrating mathematical analy-
trine, Hookerfeels compelled to ascribe such open-mindednessto him:
sis and philosophical reflection on his own work and on the work of
“Though Bohr held strongly to BC15 [the doctrine of the indispen-
other scientists (Wessels 1980; Bitbol 1996; Darrigol 1992b; Beller 1997a).
sability of classical concepts], he wouldin principle give it up [!] had
Schrédinger was accused by the orthodox of disregarding quantum
circumstances demanded it” (1991, 491).
discontinuities (chapters 2 and 4); Schrédinger, of course, did not deny
discontinuities, but he believed they could all be deduced mathemati-
Opposition, Paradigms, and Past Science
cally from a wavetheoretical scheme. He was accused of lackingthe ele-
mentary understanding that wave packets disperse (Heisenberg 1927b,
There are numerous ways to delegitimate the opposition and to dis-
73); yet he recognized the inevitable dispersion of wave packets even
credit its stand. When skillful rhetorical techniques, disguised as a dis-
before Heisenberg’s paper.
interested search fortruth, are used by powerful authority figures, their
Orthodox quantum physicists often used the argumentthat y-waves
effect is potent. It is difficult enough to produce a well-developedalter-
are not real—Schrédinger was amazed byit: “Something that influ-
native to the deeply entrenched and elaborated quantum orthodoxy;it
ences the physical behavior of something else must not in any respect
is intimidating, if not paralyzing, when all such alternatives are confi- be called less real than the somethingit influences—whatever meaning
dently ruled out by the “unbearable weight” of Bohr’s authority and by
we may give to the dangerousepithet‘real’” (1950, 110).
such scientific heroes as Heisenberg and Pauli. The G6éttingen-Copenhagen physicists discredited Schrédinger’s
 278 Chapter Thirteen
Hero Worship, Construction of Paradigms, and Opposition 279

interpretive aspirations by pointing out thatclassical waves propagate a “dogmatic”realist, as opposed to Heisenberg’s own, “nondogmatic”
in the usual three-dimensional space, rather than ina 3N-dimensional brandof realism (1958, 81). According to Heisenberg, those who ques-
space (Born 1953b, 142-43). Yet this criticism was misleading and out-
tioned the Copenhagen orthodoxy were dealing with “cracks in old
dated. Since 1927, Schrédinger had replaced his original “naive” wave bottles instead of rejoicing over the new wine” (1958, 139). The ortho-
interpretation with thatof the “second quantization approach” (accord-
dox 8ften discredited the opposition by accusing it of incompetence,
ing to this approach, one can translate any statement about N-particles
indulgence in wishful thinking, or even personality disorders. Those
in 3N-dimensional space into a statement about N-level excitations of
who did not agree with the orthodoxy were “unable to face the facts.”
the three-dimensional vacuum state). In fact, Schrédinger’s own work
Those whoallowed the possibility of hidden variables were presented
(even before the advent of wave mechanics) foreshadowed the mod-
as dreamers whohadlost touch withreality: “To hope for hiddenvari-
em concept of quantized fields (Darrigol 1986). And though he had - ablesis as ridiculousas hoping that 2 x 2. = 5”-(Heisenberg 1958, 132).
somereservations aboutthe very “abstract” characterof these “waves,” Those who seek an alternative to the Copenhagen interpretation are
Schrédinger’s preference for this scheme over the orthodox model was simply outdated “grumblers”: “This group of distinguished men. . .
unequivocal: “I believe the discrete schemeof proper frequencies of sec- maybe called philosophical objectors, or, to use a less respectful ex-
ond quantization to be powerful enough to embraceall the actually ob- pression, general grumblers” (Born 1953a, 129)."
served discontinuities in nature” (1952b, 27). . The most powerful technique for discrediting the opposition was by
Schrédinger’s attempt to substitute a wave ontology for a particle on- - identifying its views as part of an outmoded, conservative stand. For
tology was not an expressionof a naive regressive nostalgia for the old ‘this reason, the orthodox painted themselves as intellectual revolution-
way of doing physics, as the Copenhagen propaganda implied. Nor did aries. They constructed the stand of the opposition and the image of
his attempts at interpretation stem from purely philosophical presup- past science simultaneously, thus caricaturing the opposition and trivi-
positions, or “prejudices.” Since the 1930s, Schrédinger’s preference for alizing past knowledge. : ,
a wave ontology had been embeddedin his research program:to unify It is tempting, of course, to present one’s own contributions as revo-
general relativity and quantum mechanics (Riiger 1987). Following Ed- lutionary, where “revolutionary” is synonymous with novel, bold,
dington, Schrédinger hoped to deducethediscrete structure of matter original, radically new.Intensely ambitious individuals are especially
from a unification of atomic physics and cosmology. The discreteness ~’ prone to using revolutionaryrhetoric. Heisenberg hoped that his own
of a material substratum, conceivedas a continuous wavefield, follows contributions to quantum physics would “revolutionize” the twen-
from its being enclosedin a finite volume." This research program also tieth century as Copernican ideas had transformed the Renaissance
illuminated the lack of individuality and the prevalenceof holism in (1979, 21). Bohr, who wished “to realize his wishes as to the future of
the quantum mechanical domain:a single electron is inseparable from physics,” expected that his complementarity would be taughtin ele-
all others in the universe. The statistical nature of quantum theory mentary school alongside Copernicus’s ideas (interview with Bohr,
arises precisely because,illegitimately, the orthodox approach implic- AHOQP; Rosenfeld, quoted in French and Kennedy 1985, 323). Yet ori-
itly presupposes such an individuality (Ruger 1987). From the per-

2 we laEe
ginators of new theories and the judgmentof posterity do not always
spective of the unification of quantum physics and cosmology, Schré- agree about what exactly was revolutionary in their work. Planck, for
dinger’s search for a (nonclassical) model of reality was not merely example, considered his discovery of the numerical value of h, rather
natural—it was unavoidable. The search for such realistic alternative
by other physicists was aborted, in Schrédinger’s view, mainly because
12. The opposition did not spare the sharp wordseither, though these found expres-
the orthodox Copenhageninterpretation was “administered fairly early sion more often in private correspondence than in published writings. Einstein called
and authoritatively” (1950, 111). the Copenhagen philosophy a “tranquilizer” for an uncritical follower, and Schrédinger
The orthodox delegitimated the opposition by presenting it as “con- defined it as “philosophical extravaganza dictated by despair” (Landé 1965, 124). More
servative” and “dogmatic.” Heisenberg, intimating that Einstein “had eloquently, Schrédinger referred to believers in the Copenhagenphilosophy as “asses”:
difficulty” understanding the Copenhageninterpretation, labeled him “With very few exceptions (such as Einstein and Laue) all the rest of the theoretical
physicists were unadulterated asses and I was the only sanepersonleft” (Schrédinger to
Synge, quoted in Moore 1989, 472).
Il. The discreteness is deduced because the vibratingfield, enclosedin
a finite volume, 13. In classical times one used the word imutatio for novelty. In the sixteenth and sev-
can have only discrete proper modes. :
enteenth centuries one used the word “new” rather than “revolutionary.”
 280 Chapter Thirteen
Hero Worship, Construction of Paradigms, and Opposition 281
than discreteness of the quantum ofaction,
as constituting a “revo- position and the construction ofthe scientific past. It was because of
lution.”
Exactly where one chooses to demarcate betwe Schrédinger’s challenge and Einstein’s critique in the late 1920s and
en the old and the new early 1930s that the Gottingen-Copenhagen physicists chose acausality
depends on the local theoretical and
sociopolitical context. While in
the 1930s Heisenberg presented relativity theory and indeterminism asthe focal points of their emerging quantum para-
and
quantum theor
as together opening a newera inphysics digm (rather than the more récently proposed nonlocality).
(the quantum overthrow of
Kantian causality is adirect continuation ofthe Asthis new paradigm emerged,its founders constructed a profile of
Einsteinian overthrow
of Kantian space-time), both Einstein and relativity the opposition and a description of past science simultaneously. The
disappeared from ideas of the opposition were projected as most characteristic of the
_ Heisenberg’s speeches and writings during the Third Reich (see
the overthrown past—in this way opponents were automatically presented
collection of his essays, Heisenberg 1979). In the
postwar years, when as conservatives; disposing of the old and discrediting the opposition
Einstein was perceived as the most prominent
opponentof the Ce en-
hagen interpretation, the line again was drawn went hand in hand. The opposition became simpleminded and reac-
dilferently: —selalivity,
whichstill preserves the notions of causality and tionary; the past became monolithic. The diversity, ingenuity, fluidity,
of an invariant objec- and epistemological resilience of past science was thus forced into a few
tive reality, belon gs to the prerevolutionary past—it is quan
that opened a genuinely newage.
tum theor r rigid, simplistic categories. In this wayclassical physics became uni-
. >
Whether one chooses to presenta contribution as- formly deterministic. Yet probabilistic ideas were introduced into clas-
what feature one-singles out as the “most revolu tionar y, and sical physics as earlyas the beginningof the nineteenth century (see the
revolutionary,” also depends
‘

on local theoretical and sociopolitical circumstances, Heise papers in Kriiger, Gingerenzer, and Morgan 1987). Statistical methods
interpretation paper (1925) is usually accept nberg’s re- were the mainstream in quantum physics since Planck's and Einstein’s
ed today as the basis of workat the turn of the century, before the new quantum theory. This
the new quantum theory and as the inauguration of the
conceptual was the reason Born did notinitially regard his probabilistic interpre-
revolution in physics. Yet initially neither Heise
nberg nor Born and Jor-
dan, who extended Heisenberg’s ideas (Born and Jorda tation of the wave function as signifying a revolutionary departure,
n 1925a: Born the way it was later construed by the Copenhagen physicists (interview
Heisenberg, and Jordan 1926), presented the
new quantum theor y as with Born, AHQP).
revolutionary. Quite the contrary—because ofthe highl
y abstract char- Norwereclassical physicists naive about the possibility of exact pre-
acter of the new quantum formalism, they preferred to empha
sizeits dictability. Since Poincaré’s work at the turn of the century, physicists
connection with the past and with cherished
classical ideas (the iden- were aware of the existence of chaotic systems—nonlinear dynamical
tical form of the canonical equations, the identity of
the perturbation systems with sensitive dependenceoninitial conditions.’ Born himself
methods, energy conservation). No indete
rministic conclusions were
initially deduced from the formalism (chapter 2). The rhetor was cognizant of this phenomenon—he wrote a paper with an un-
ic of
a re- equivocal title: “Is Classical Mechanics in Fact Deterministic?” (Born
turn to classical purity and beauty was abrupt
ly silenc ed when some-
thingstill more classi cally pure and beautiful appea 1955a). The prominent orthodox argumentthat quantum physics gives
red—Schrédinger’s
competitive wave theory. I have argued that Heise up causality in principle, while classical physics gave it up only in prac-
nberg and other or- tice, wasitself crystallized in this tendentious construction of the past.
thodox physicists developed the arguments for a revol
utionary over- It is the very meaningfulness of this distinction between indeterminism
throwof causality as a response to Schrédinger’
s competing theory of
quantum mechanics—a continuous,
causal alternative. This example
in principle and indeterminism in practice that chaos theorists chal-
indicates howthe choice is made of which
ideas to label Aeonservalive 0
Tenge today.
This analysis also suggests why some elements, rather In the simultaneous construction of the opposition and the past, the
are chosen to serve as foundational pillars than others opponents—such as Einstein, Schrodinger, Bohm, and Landé—were
for a new paradigm. It also
points out the intimate connection between a challe presented as conservatives, unable to digest the revolutionary novelties.
nge from the op- Ihave not been able to find any commoncriteria by which their stand
14. His idea was that h, together with other fundamental constants—e (electr was characterized as conservative—their conservatism consisting
charge), k (Boltzmann’s constant, andc (the speed of light)—wou
ld provide absolute ies
versal units of measurement, independentofthe2 “accident” of humanlife on Earth . Fo
For
details, see Klein (1977). 15. An error in the determination of the initial conditions quickly becomes large
enough to preclude the possibility of predicting the futurestate of the system.
 282 Chapter Thirteen
SOHA AS HAR XG

Hero Worship, Constructionof Paradigms, and Opposition 283

merely in their disagreement with the orthodoxy. We are all familiar
with the caricatures of the opponents—the image of aging Einstein kinds of activity require, each of them, suchelaborated special training
stubbornly mumbling “God does not play dice.” Weare also familiar andskill, that they are seldom commendedbythe sameperson” (Schré-
with Galileo’s construal of the fictional Peripatetic Simplicio. As Gali- dinger 1954, 124).
leo’s brilliant caricature distorts and disguises the complexity and resil- he objectification of measurements, the intersubjective agreement
ience of Aristotelian thought (Schmitt 1983), so too the Géttingen- about the results of experiments, is primarily based on geometric and
Copenhagen physicists trivialized and distorted the ideas of Einstein kinematic statements. For it is in the assessment of a geometric coinci-
Schrodinger, and Bohm.Recent scholarship, notably Arthur Fine’s and dence between the marks on the scale and the pointerthat the discrep-
Don Howard's work, reveals that Einstein was neither a simpleminded ancy between thejudgments of two different observersis the least pos-
determinist (he did not hope for a completion of quantum physics by sible. Inthis sense, empirical statements aboutgeometric and kinematic
hidden variables, but for a radically new theory that would subsum relations clearly have a privileged epistemological status. Yet this state
e of affairs contributes to the fallacious naive realist view that “some-
the current quantum physics—Fine 1986), nor was he an
unsophis- thingis distributed in space in a definite arrangement and well defined
ticated realist—in fact, his philosophy had prominent conventional-
ist strands (Howard 1990, 1993). Far from holding a correspondence
order,” this “something” is changing according toobjective laws, and
theoryof truth, Einstein considered the concept of independent physi- “this changing something” constitutes so-called objectivereality (Schré-
cal reality meaningless (Fine 1986; Beller, forthcoming). Nor was Ein- dinger 1954, 144-45).
stein a dogmatic adherent to causal, continuousfield theories—there ’ Positing this objective reality leads, according to Schrédinger, to a
epsaaptensetator nesting: te

was “another Einstein,” and the debate between Einstein’s two conflict
- very awkward philosophical situation. Since “secondary” sensory ex-
- periences (sensations of color, taste, smell, sound) are removed from
ing voices lasted to the end ofhis life (Stachel 1993).
“objective reality,” one invents a new realm for them—the mind(!),
Recentscholarship has also refuted the picture of‘Schrédinger as an
Bs

adherent to realism. As did Einstein, so too did Schrédinger deny the forgetting that “all that we have been talking abouttill nowis also in
meaningfulness of the concept of independent reality, regarding the the mind and nowhereelse.” This division creates philosophical pseu-
doproblems, such as how mind and matter act on each other, and the
idea ofreality rather as a regulative construct (Ben-Menahem 1992:
Bitbol 1996; Beller, forthcoming). like (Schrodinger 1954, 145).
ce 4 Schrédinger’s draft for the James lectures (1954) allows us to form a
Schrédinger’s defense of a comprehensive wave ontology, as well as
the persistent Copenhagencaricature of Schrédinger’s position, has re- clearer view of his aversion to the strong positivist elements of the Co-
sulted in an image of Schrédinger as a conservative, simpleminded penhageninterpretation—an aversionthat the Gottingen-Copenhagen
classical realist, unwilling and unable to sacrifice traditional] concepts
physicists mistakenly characterized as a form of simplemindedrealism.
and accept new ones. Yet an analysis of Schrédinger’s writings reveals An analysis of these lectures reveals why Schrédinger fundamentally
instead a very sophisticated position, along neo-Kantianlines: the con-
rejected another dichotomy—Bohr’s notorious division between the
micro- and macrodomains, between classical “reality” that is close to
cept of reality “as such,” as it objectively exists independentof all hu-
manobservers,is indefensible, if not downright meaningless. Similarly, commonsense and the quantum mechanical “abstract formalism’ that
Schrédinger fully understood that the correspondence theory of truth _ is a tool for description and prediction of measurementresults. Schré-
can hardly be sustained.Still, the coriceptofreality, held Schrédinger dinger rejected this division not because he extended the naive realism
is as indispensable in scienceas it is in everydaylife. There is no icp of the macrodomain into the microdomain. Quite the opposite: he de-
tinction in principle between a layman’s anda scientist's nied to the macrodomain realistic statusthat the Copenhagen philoso-
conception of
reality—both are regulative constructs, indispensable for phy regardedas a self-evidentfact.
mental (and Schrédinger was fully aware of the pitfalls, deficiencies, and ideo-
physical) activity.
Schrédinger, no fan of duplications and divisions, logical powerof positivism. He called the legitimating positivist stra-
dismissed those
dichotomies on which representative realism rests—the dichotom tagems of the Copenhagen orthodoxy “unfair subterfuge” (1952d, 83),
ybe- a “supreme protector” (1935, 157). The positivist approach entrenches
tweenprimary and secondary qualities, between mind and matter,
be- . the basic assumptions of the Copenhageninterpretation and endows
tween theory and experiment. Thedivision into experim
ental and eeu
retical physicsis artificial—it is “mostly caused bythefact that the controversial assertions with an aura of inevitability. The Copenhagen
two claim that quantum mechanics is complete is among such assertions,
 284 Chapter Thirteen
Hero Worship, Construction of Paradigms, and Opposition
285
prohibiting further questioning. Complete
ness seems to arise as a
straightforward deduction from positivism: “It must be impo Philosophically sophisticated and technically ingenious are
also Boh-
add onto it additional correct statements, without ssible to mian alternatives to the Copenhageninterpretation (Cushing 1994a;
otherwise changing
it; else one would not have the right to call meani Bohm and Hiley 1993). And AnthonyValentini has recently argued that
ngless all questions
extending beyon dit” (Schrédinger 1935, 159). de Broglie’s scientific contributions have also been misconstrued in a
Schrédinger was no less a philosophical “oppo grotesque way—de Broglie was not somecrankyoutsider, or second-
rtunist” than his rate theorist, but a deep thinker and an outstanding physicist, who in
Gottingen-Copenhagen opponents (compare chapt
er 3). He did not shy fact as early as 1927 (twenty-five years before Bohm)hadderived all the
away from positivist arzuments when they suite
d his purpo ses. In fact
it seems that Schrédingerwas nolegsskillful in posit correct “pilot wave” dynamics for a multiparticle system.*
ivist analysis when As to the claim that the opponents did not have a deep working
criticizing the orthodox camp than the G6ttinge
n-Copenhagenphysi-
cists were when fending off the challengesof the Opposition knowledge of quantum physics, anyone who is even superficially fa-
. In such miliar with Einstein’s and Schrédinger’s publications immediately sees
criticisms, Schrodinger often presupposed
the verificationist meaning
of quan
tum formulas: the uncertainty relations, for examp howridiculousit is. It was the opposition—Einstein and Schrédinger—
le, are not whoin the mid-1930s discovered and mathematically elaborated the
merely limits on the possible measurement value
s of Physical vari-
ables—uncertainty restricts the very definability of the basic nonseparability of quantum systems (Einstein, Podolsky, and
concepts used Rosen 1935; Schrodinger 1935). It was the orthodoxy, as I have azgued,
(see his criticism in chapter6). Similarly, Schrodinge
r used positivist
ey t

arguments when “deconstructing” the conceptof whodiffused these arguments byoperational stratagems, preventing
es

a particle: “We are

serious exploration of nonseparability until Bell’s seminal work. Ein-
st

not experimenting with single particles, any more

than we can raise
m a
Peer pe

Ichthyosauria in the zoo. We are scrutinizing recor stein’s characterization of Bohr as a “Talmudic philosopher” referred
st

ds of events long
after they have happened” (Schrédinger 1952a, 240). precisely to Bohr’s circumventing, rather than directly confronting,
As Born noted the most fundamental problem of quantum theory. This deep physical
Ti

shrew dly, refusing to go from an experimental event

to an underlying challenge was met with the rhetoric of “sacrifice.” ” The “sacrifice” pri-
substratum (particle matter) is a positivist argument(19
53b, 144). .
ESS
Sp eee

A letter from Schrédingerto Eddington (written in 1940, marily meantelimination of the opposition’s ideas.
reprinted in
Pe

Bitbol 1995, 121-22) reveals that Schrédinger indee

d considered him-
self simultaneously an heir to Machean positivism and 16. This assessmentof de Broglie’s work is opposedto the received view that de Broglie
a follower of developed only a primitive case, treating y as someclassicalfield in three-dimensional
Boltzmann’s descriptive tradition. The approaches of
Boltzmann and space. Valentini also challenges the acceptedstory thatat the fifth Solvay conference de
Mach,atfirst sight irreconcilable, werein
fact directed toward the same Broglie was unable to reply to Pauli’s penetrating criticism and he therefore abandoned
“goal: avoiding doubtful presuppositions, his efforts for an alternative to the Copenhagen orthodoxy. Iam grateful to Valentini who
excluding contradictory as-
sumptions, clearing obstacles on the path to truth. While shared with me his reinterpretation of de Broglie’s work in private communication before
“Boltzmann ’s
idea consisted in forming absolutely clear, almos publication. This subject is treated in Valentini (1998). It must be noted that Valentini‘s
t naively clear and de- interpretation is controversial. Scholars disagree about wherede Broglie stood conceptu-
tailed ‘pictures’— mainlyin orderto be quite sure
of avoiding contradic- ally in 1927, and about the success with which he countered Pauli’s objectionsat the Sol-
tory assumptions”—Mach “was most anxious not
to contaminate this vay conference (James Cushing, private communication).
absolute reliable timber [an economical summary 17. Bohr, Heisenberg, and Born talked repeatedly about a “sacrifice” of the old. The
of the observed facts]
with any other one of a more doubtful origin”(Bitbol following statement is typical: “All progress has been achieved bysacrifice” (Heisenberg
1995, 121). Thus
Schrédinger wrote to Eddington that he could hardly 1979, 27). The rhetoric of sacrifice is noted in Heilbron (1987, 219).
be impressed by
the “brave new world” of the Copenhagenpositivism
—a rather crude
and naive version of an approach he wasalways famili
ar with.While
one may and even must use pictures, one has to do
so with one’s eyes
opento their limitations, revealed by an analysis
of the experimental
possibilities (Bitbol 1995, 122). It is the relati
ve weight of positivist and
model-descriptive elements in Schrédinger’s
arguments that changed,
one can conclude, according to the theore
tical challenges that he en-
countered.
 CHAPTER 14
EE

Dialogues or Paradigms?

Physics gets authorityfromideasit propagates but never obeys in actual

research, methodologists play the role of publicity agents whom
- - _ physicists hire to praise their results.
Paul Feyerabend 1993, 261

Thereisa crack in everything. Thats how the light goes in.

Leonard Cohen, “The Future”

Introduction

Thomas Kuhn’s Structure of Scientific Revolutions (1970, hereafter re-

ferred to as Structure) is the best known and mostinfluential critical
modelof science in our time. Yet the central concept of Structure—that
of a paradigm—is incompatible with the notionof criticism. Criticism
presupposesthe existence of a contemporary alternative—the concept
of a paradigm denies it. By definition, a paradigm excludes any diver-
sity of opinion or open-minded discussion. It legitimates dogmatism
andsilences the opposition. Not surprisingly, Kuhn andsociologists of
science following in Kuhn’s footsteps talked in terms of “deviation,”
“impermissible aberration” (Kuhn 1970, 209), or “deviance”rather than
legitimate disagreement.
In this chapter I argue that the notion of a paradigm has not only
clear totalitarian implications but also dogmatic ideological roots. I
argue that close historical links exist between the notion of incommen-
surable paradigmsandthe ideologyof the Copenhagen dogma.In par-
ticular, I disclose the importance of Bohm’s alternative to the Copenha-
gen interpretation in the emergence Of a post-positivist philosophyof
science.
The notion of a paradigm is also historically inaccurate. Historical re-
search on different episodes in the history of science reveals not incom-
mensurability but rather openness, selective borrowing, and commu-
nication. This openness and communication suggest that a dialogical
 288 Chapter Fourteen
Dialogues or Paradigms? 289
approach is also appropriate to describe the growth of ‘
fields other than twentieth-century physics. 2 Senewaledge in quacy, but the use of the concept of a closed theory that concerns us
here. A paradigmatic case of a correct closed theory is, according to
Heisenberg’s “Closed Theories” and Kuhnian “Paradigms” Heisenberg, quantum mechanics. Under the Copenhageninterpreta-
tion, quantummechanics can neither be modified by small changes nor
The Kuhnian notion of a paradigm supplemented by Ridden variables. Therefore, argued Heisenberg (as
is peculiarly reminiscent of Hei
berg’s notion .of a “closed theory”
(Heisenberg 1948, 1958 1979).Ac. well as Bohr and Pauli), the attempts by the opposition to challenge
cording to Heisenberg, progress in science quantum theory and its orthodox interpretationare futile (chapter 9). It
occurs throu ha s enc
of closed theories. A closed theoryis true was usually in the context of combating criticisms of quantum theory
in the limitedtoon feofita
validity andis not open to modification that Heisenberg invoked the notion of a closed theory. The notion of
.1 A closed theor a ot be
improved by small changes; it can only a closed theory implies that the advance from an old theory to anew
be replaced by somethi aa 7
tially different. The notion of a closed theo one demandsanintellectual jump (very much like the Kuhnian Gestalt
ry, like the notion of:Kuh.
ian paradigm,has a strong holistic aspe switch): “From the point of view of quantum theory . . . Newtonian
ct: “The connection betwe n fe
different concepts in the systemis so close that one coul mechanics cannot be improved; it can only be replaced by something
d dnevally not
change any one of the concepts without destroyi essentially different... . This realization can preserve us from [the]
(Hei
ng the whol e s ser mistake, not always avoided in the past, of attempting to force new
senberg 1958, 94).2 From timeto time
, physicists disc
over Tee whe €
they have held to be an unshakably true ' fields of experience into an outmoded unsuitable structure of con-
theory (classical ph sics oy
has only a limited range of validity. How can onetalk in ach cepts. ... We shall serve the future best by . . . easing the wayfor the
stances of exact science at all? (Heisenberg som newly won methods of thought rather than by combating them” (Hei-
1979, 44). We can save the
notion of an. exact science bysacrificin senberg 1979, 26). é
g the notion of unity. b reli :
quishing the Einsteinian goal of an ultimate Neither the philosophy nor the history with which Heisenberg sup-
unified theory. dence a ported his notion of a closed theory is convincing. Quantum mechan-
collection of independent “islands,” of
different doriains
, which a
not be united into a commonstructure. With ics can be supplemented by hiddenvariables, albeit nonlocal ones, de-
in the domain ofa clossd spite von Neumann’s so-called impossibility proof (Bohm 1952; Bub
theo ry we can useits basic concepts with
complete confidence.
How d
we knowthat in the domain of a closed theory its 1997). Noris it true that quantum mechanics cannot be modified by
exact? Heis
basi c conc e t ° small changes—a notable example is the Ghirardi-Rimini-Weber the-
enberg’s answeris a careless one: “Wh
unhesitantly, they are exact” en we use me ots ory (1986), which circumvents the measurementproblem by modifying
(1979, 44). Heisenberg, of cour
provide us with any independent guid se sae : Schrédinger’s equation. Nor can we reasonably holdthatclassical phys-
elines about when we can is

vba
concepts “unhesitantly” unless we already ics is a closed theory, given the host of new, nonclassical phenomena
believe that they are cor :
i and exact—a characteristic circularity of Heisenberg’s argu
ments ~
‘disclosed in its domain by research in chaostheory.
Heisenberg’s idea of a closed theory, conceived in a context of con-
S There are other obvious weaknesses
in Heisenberg’s notion of
closed theory—he provides no argument for identify frontation, is an ideologicai notion, aimed at a defense of quantum
ing the corr fe
ee

te ness of a theory with its completeness, nor does he indicate any c theory andits orthodox interpretation.It is a powerfultoolof legitima-
iter ‘ tion in the rhetoric of the irrefutability and finality of the quantum
that determine the limits of applicability of a closed theory (a a
characteristic circularity: “classical phys theory. I argued in chapter 9 that the central Copenhagenpillars, such
concepts can be applied’’—1979, 23).Y ics extends just as faras‘ite as the doctrine of the inevitability of classical concepts and operation-
etit is not Philosophical inade.
alism, are powerful social strategies of legitimation, disguised as ob-
1, “Scientific
L systems must be-complete in orderto be correct” jective philosophical arguments. The notion of a closed theory is also
ct” (H i
2. This mass has recently been emphasized in Kuhn (1987). Cetsenbeng 1979,/29)
ingeniously constructed to argue for the inevitability and finality of
. 3. Heisenberg’s notion
t of a “closed theorv
‘ ” has received very quantum theory and to encourage uncritical commitmentby its practi-
tion: It was treated briefly by Chevalley rylittl
li e scholarly RS
(1988), who pointed out the sail
sno tioners, very much like the Kuhnian paradigm.
Heisenberg’s notion and the conception of a physical theory
formalism, along the lines of logical positivism. as a Partially inte rpreted
on There are, of course, differences between Kuhnian paradigms and
It was used imaginatively by
(1992a, 1992b)in his analysis of stable domainsin theoretical and
Hacking Heisenberg’s closed theories. Kuhn’s paradigms, when replaced, are
laboratory science. forever buried in history. Heisenberg’s closed theories are correct for
 290 Chapter Fourteen
Dialogues or Paradigms? 291
all time. They serve as indestructible memorials to
the greatn
essof thej
discoverers.* The intensely ambitious commensurability. Revolution need not be a violent, Gestalt switch of
Heisenberg, not surprisin iy,
turns out to be the author of such an eternally meanings, practices, and allegiances, as Kuhn’s own Copernican Revolu-
true theory ‘As Wel -
sacker putit, not without envy: “The concept tion (1957), written before Structure, so impressively demonstrates.
originated in a reflection
on what Heisenberg himself had achie
ved twenty years earlier when
Whatweseethere is a gradual, sometimesirresistibly rational, tran-
-mha d enoughto lay the foundation ofthe latest sition from the “closed world to the infinite universe,” an almost “in-
closed theory” evitable” incorporation of atomism into this process, and a constant
Despite this difference, it is the similarities “essential tension” between tradition and innovation. As Robert West-
between closed theori
and paradigms thatare significant. Both resist man (1994) indicates in his valuable essay, there is neither incommen-
improvement b all
changes. Both are holistic. Both demandthat adva surabilitynor theory-ladenness of observation in The Copernican Revo-
neces come by wear
wayo f
jumps to a qualitatively newintellectual exper littion,.though oneis often.tempted to read them back into the book
ience. Both imply “i
commensurability.” And both require dogma underthe influenceofStructure. The author of The Copernican Revolution
tic commitment and dis.
couragecriticism. No wonder, when Heisenberg appearsto be a faithful disciple of Alexandre Koyré: Kuhn’s book is an
read Kuhn’s book h outstanding conceptual history emphasizing theory at the expense of
foundit all too familiar: “Alright, but the
new paradigm is a newcl d
theory”(quoted in Weizsacker 1987, 281). observation, analyzing the influence of wide philosophical and cultural
mers
Are these similarities coincidental? Or can currents at the expenseof local everyday practice, and carefully laying
we find close historical
links between the philosophy of Structure and out the fabric and inner logic of the evolution of concepts.* And while
the ideology of the Co- Kuhn, like Koyré; perceives revolution as the destruction of an old
penhagenorthodoxy? And, in particular, can
we uncover the roots of world and its replacement by a largely incompatible new world, noth-
Kuhn’s novel and controversial notion of
incommensurability?
ing in Kuhn’s analysis foreshadowsthe radicalization of the conceptof
Where Did Kuhnian Incommensurabili revolution introduced by Structure.”
ty Come From?
One can argue, perhaps, that Kuhn’s insightful and sympathetic de-
Thomas Kuhn himself considered the notio scription of rejected. knowledge, his presentation of the Copernican
n of incommensurabilit th
“central innovation introduced by the book [Siructure - and Ptolemaic astronomies as having the same evidential basis, his in-
pars ALR aa oa eG a TE

]” (1993 315)
Tt sistence that in the end Copernicus’s system suffered from the same
is this notion thatis tied to the most
radical tenets of a ral
tivism, lack of rationality in theory choic degree of inaccuracy, inconsistency, and complexity as the Ptolemaic
e, theory-ladenness of ob.
servation, changes in fundamental perception model, imply a rejection of the notion of truth and the introduction of
s and intuitions follow- relativism.. Yet Kuhn is far from acknowledging such a possibility:
ing the acceptance of anew paradigm,lack
of communication between
representa
tives of different paradigms, Gesta “Each newscientific theory preserves a hard core of the knowledge
lt conversion from one
paradigm to another. Over the years, provided by its predecessor and addsto it’’ (1957, 3).§ At most, Kuhn’s
trying to escape therelativism of
Structure, Kuhn developed a much milder, taxon
omic notion ofin- 6. A good exampleis Kuhn’s analysis of the gradual developmentof the concept of
commensurability, identifying it with linguistic
untranslatability. Kuhn “sravitational force” (1957, chap. 7).
emerged from his search feeling more stron
gly than ever thatincom. 7. Compare Kuhn and Koyré: “As science: progresses, its concepts are repeatedly de-
44

mensurability must be a component of stroyed andreplaced. .. . Like Aristotelianism before it, Newtonianism at last evolved...
any historical, develo mental
or evolutionary view of scientific knowledg problems and research techniques which could not be reconciled with the world view
e” (1991 3). It isihe i i that produced them.Forhalf a century we have been in the midst of the resulting con-
ial ceaof incommensurability
acbacis : . into S tructure andits histotical
vee Li TAT

a ee ee asttn ‘ . ceptual revolution” (Kuhn 1957, 265). Similarly, Koyré: “The founders of modem science
and ideological consequencesthat had to destroy one world and to replace it by another. They had to reshape the frame-
workofour intellectitself, to restate and_reform its concepts” (1943, 405). Yet this process
The notion of a scientific revolution, is the opposite of a Gestalt switch: “The spiritual change did not occur in a sudden muta-
of a radical and com rehensi
change, need not contain the controvers tion. . . . revolutions too need time for their accomplishments. . . . the heavenly spheres
ial and pregnant notion ofin did not disappearat once. ...the world-bubble grewand swelled before bursting” (Koyré
4. As Galileo'
, s Medice
: an planets were conceived b y Galileo 1957, viii). It is this evolutionary conception of revolutionary change that underlies
i as ani i
morial to Prince Cosimo and to himself; see Kuhn's Copernican Revolution.
Westfall (1985) and Bia: ‘oliae
5. For an excellent comparison between Kuhn’s as
initial . ; :4 8. At the very least, this “hard core” consists of observational results—wehave not
surability, see Maudlin (1997). val and recent notions ofincommen- even a hintat the radical analysis of “seeing’’ provided in Structure, which maintains the
theory-ladennessof observation.
 292 Chapter Fourteen
Dialogues or Paradigms? 293
analysis is compatible with, and perhaps influ
enced by, anotherthesis Kuhn’s “encounter with incommensurability” and his discovery and
that of the “underdetermination of theory
by data.” "Yet this th ae elaboration of the concept of a paradigm must have been connected.
though undermining the notion of a single
truth, does not exclude -- Kuhn's reminiscences locate the beginning of his work on Structure in
the sameradical way as theory-ladenness of obser
vation does: a 1958-59. It was during this year that, after having “great trouble” de-
underdetermination does not imply the imposs
ibility of commiani ~ veloping his ideas (Kuhn 1977, xviii), he “finally had a breakthrough”
tion. As long as communication takes place,
one canat least attem ie 5 (Horgan 1991, 14) achieved by finding the “missing element” (Kuhn
single outconsensual truth by appeal to shared
epistemic virtues The 1977, xix)—the concept of a paradigm. It must have been at the same
nmuscommunication inherent in the notion of the
incommensurabilit f time that Kuhn came upon the notion of incommensurability: “My own
paradigms leaves no such option. Kuhn, indeed,
presents an ast oe encounter with incommensurability was the first step on the road to
mer’s choice of the Copernican system overthe Ptole
maic as duet ened Structure’ (Kuhn 1993, 314-15). Where could Kuhn have met with this
Platonic considerations, harmony and mathematic
al beauty. “ee notion?
Onevor the most fascinating aspects of The Coper
nican Revolution is Another-book appeared before Structure, a book that undertook a
Kuhn’ s penetrating description of the “inter
face” between tradition radical analysis of the notion of seeing, of the theory-ladenness of ob-
and innovation.” The delicate balance between tradit
ion and innova- servation, of Gestalt experiences in perception, of the revolutionaryre-
tion, conservatism and radicalism, “convergent” and
“divergent” think- structuring of an old pattern into a new one, with the resulting loss of
ing, commitment and open-mindednessconstitutes
the focus of Kuhn’s the ability to communicate. In this book the word “paradigm”appears
analysis in his “Essential Tension” (1 959), an essay
written when he be- on the first page.It is, of course, Norwood Hanson’s Patterns of Discov-
gan his work on Structure. Kuhn’s struggles with the
issue of conserva- ery, which appearedin 1958— an important book that would betotally
tism versus innovation reveal an acute historical
sense of the continuit eclipsed by Structure.
of changerather than an uncritical commitment to
the idea, of steel Hanson’s book opens’ with a fictive encounter between a pro-
frameworks. Yet the “essential tension” swiftly devel
oped in Kuhn’s Copernican Kepler and an anti-Copernican Brahe and with the surpris-_
hands (1963), into a celebration of “dogma”in scient
ific research “Pre. ing and explosive question: “Do Kepler and Tychosee the samething
conception andresistance to innovation” became, for
Kuhn hallmarks in the east at dawn?” (1958, 5; compare with Kuhn'sfictive speech by a
of the scientific identity necessary conditionsfo
r scientific creativity, Copernican convert—1970, 115). By discussing different organizations ~
indispensable aspects of scientific education.”
The forthcomin Struc. of visual data exemplified in Gestalt phenomena (a bird versus an an-
ture disclosed a deep conceptual connection
between ee telope, an old Parisienne versus a young woman4 la Toulouse-Lautrec,
bility and the notion-of a paradigm: incommensu
rability excludes th a perspex cube viewed from below versus above—suchcubesdecorate
possibility of being suspended between two differ
ent. incom atible the cover of Structure), Hanson prepared the groundfor his penetrating
worlds, of creatively participating in both,of susta
iningfor lon: a rely conclusion that “seeing is a ‘theory-laden’ undertaking” (“observation
ative tension between the old and the new. Such work
is eee only of X is shapedby prior knowledge of X”; Hanson 1958, 19). Tycho and
during a-short period of crisis, disarray, incons
istency. iwsiticaan. Kepler do notsee the same thing and merely interpretit differently; they
rability logically dictates total, unquestioning, dogma
tic commitment do have in fact incompatible visual experiences: Tycho sees a mobile
we
9. aseie
Se i ' analysis
i of Copernicus’s
icus’: attempts to “justify”
sun, Kepler sees a static sun. “Seeing as” and “seeing that’ become in-
the earth’s mo-
distinguishable. It is because there is no pure observation language, it
10. “Preconception and resistance se
ion i
is because the data themselves are different for Tycho and Kepler, that
scientific development” (Kuhn 1963, ort: aia on “ saat they live in incompatible worlds. Weare familiar with these arguments,
Stramiental in making the sciences the most consistently
revolutionary of all h piel of course, from Kuhn’s “Revolutions as Changes of World View,”
tivities 5 “the commitment is actually constitutive
mains
of research”; “scientific edueatinn vo.
a relatively dogmatic initiation into a pre-established chapter 10 of Structure—argumentsthatare the basis of incommensur-
problem-solvin
that the studentis neither invited nor equipped to evaluate” (349). This essay = ee ability.
- een
before 1963, and probably before 1962, for Kuhn informs
his readers: “The eae ne Hanson’s work, including his extensive discussion of the connection
oped in this paper have been abstracted... from... between perception and language, “pictures” and “sentences,” “facts”
my forthcomin eae ga
Structure ofScientific Revolutions, which will be published in 1962
Press”(1963, 347 n. 1). byU: . er me and “expressions,” in the spirit of Wittgenstein, deserves closer atten-
sleevesseeeune4 tion. So doeshis possible impact on the historiographical revolutionof
 294 Chapter Fourteen
Dialogues or Paradigms? 295
the 1970s in general, and on Kuhn’s work in particular. Yet on the more
narrow, though absolutely central, issue of incommensurability: material that did notfit his aims: “Any argument not applicable to mi-
the crophysics has been held generally suspect; conversely, arguments have
tight connection between Hanson’s and Kuhn’s ideas
is undeniable. The been regardedasestablishedif they help one to understandthe concep-
evidence is not simplythe similarities between the two texts,
but Kuhn’s tual basis of elementary particle theory” (1958, 1-3).
ownreferences to Hanson’s work. While discussing the oy. : cae : wy:
Gestalt changes While discussing historical cases of incommensurability—between
of paradigm during revolutions, Kuhn mentions Hanson, who “has
used Gestalt demonstrations to elaborate Tycho and Kepler, Descartes and Beckman, Mach and Hertz—Hanson
some of the same conse-
quences of scientific belief that concern me here” referred repeatedly to thesituation in quantum physics as the prime
(1970, 113). Another example of a newconceptualpattern’s incompatibility (logical discon-
essay (written soon after Structure appeared), in which Kuhndiscussed
the “revolutionary reconceptualization” that tinuity) with the old? >
permits data “to be seen’
in a newway,” again refers the reader to Hanson’s works Whatis, then, the content ofHanson’s final chapter? The bulkof the
.” Kuhn‘s ° chapter is devoted to description and approbation of the Copenhagen
use of incommensurability goes, of course, beyond
Hanson‘s—itis the dogmaand argues the impossibility of an alternative to the orthodox
powerful synthesis of philosophical analysis, sociol
ogical incorporation interpretation. Hanson’s writing spirals into opacity’? as he approaches
of the notion ofa scientific community, and rich histor
ical knowledge the issues of the holism of paradigms and incommensurability.
and insight that makes Structure a masterpiece. Yet
the basic similarity Hanson's discussion was very close to Heisenberg’s writings, some-
bet'veen Hanson's and Kuhn’s notions of incommensu
rability leads di- times embarrassingly so. He dutifully repeated Heisenberg’s arguments
rectly to the following question: Where did Hanson get this
controver- about the impossibility of visualization in the quantum domain, about
sial, if not bizarre idea that Tycho and Keplerlived in
incompatible wave-particle duality, and about the inevitability of acausality, fol-
worlds and were unable to communicate? As we know,
theylived (at lowing from Heisenberg’s uncertainty relations. As for Heisenberg, so
least for a while) on the grounds of Tycho’s castle and
did, in fact, argue for Hansonthe uncertaintyrelations are the cornerstone of the whole
constantly, loudly, and passionately. What, then, could have
suggested quantum theory—their violation is a “conceptual impossibility.” This
to Hanson theidea of incommensurability? What were
Hanson’s direct does not mean that the uncertainty principle is a tautology, or defini-
sources and aims?
tion: “Had nature beenotherthanit is. . . the principle might never
have been formulatedat all” (1958, 136).
Hanson's Incommensurability and the Copenhagen Dogma
Hanson used wave-particle duality to deduce the “conceptual impos-
As Hansoninformedhis readers in the introduction to Patter sibility” of the simultaneous determination of the positions and mo-
nsof Dis- menta of particles. Note that in Bohm’s theory, position and momentum
covery, “the approach and method of this essay is unusua
l.” Disap- can be defined simultaneously, though they cannot be simultaneously
pointed with the inadequacy of discussions concerning
the nature of measured—Bohm’ss version of quantum mechanics constitutes the best
genera l philosophical issues such as the status of observ
ational facts disproof of such “impossibility” arguments. Hanson relied uncritically
andcausality, Hanson decided to use current atomic theory as

J WieAh
“the lens on the usual Copenhagen assumptions (those that the orthodox pre-
through whichthese perennial philosophical proble
mswill be viewed.” sented as indisputable), such as the indispensability of classical con-
All discussions in his book, Hanson informed
readers, “are written cepts in a quantum theoretical interpretation. This resulted in Han-
with the final chapter in mind”—the chapter in which
he discussed son’s repetition of the usual Copenhagen rhetorical strategy, presenting
the lessons of the quantum revolution. Hanson’s
credo wasnotsimply
Whiggish. In an outspoken way, he refused
to cope with anyhistorical
12. “To say that Tycho and Kepler, Simplicius and Galileo, Hooke and Newton,Priest-
11. “N. R. Hanson hasalready argued that whatscientis ley and Lavoisier, ... Heisenberg and Bohm all make the same observationsbut use them
ts see dependsupon their prio
beliefs andtraining, and much evidence on this differently is too easy. It does not explain controversy in research science” (Hanson 1958,
point will be found in the last referen :
cited in note 31” (Kuhn 1977, 263 n. 33). Similarly, “N. R. Hanso 19): “It is the sense in which Tycho and Kepler do not observe the same thing which must
n... has used 3
demonstrations to elaborate some of the same
consequences ofscientific belief an
be graspedif one is to understand disagreements within microphysics” (18). The differ-
cernme here” (Kuhn 1970, 113). There are other references to Hanson’s workin Séru ence between Descartes’s and Beckman’s idea of motionis “like Einstein, de Broglie, Bohm
as well as in almostall the other essays Kuhn 7 and Jeffreys on the one hand, and Heisenberg, Dirac, Pauli and Bethe on the other,all
wroteat the time he was workin a fe
ture; see the paperscollecte
d in Kuhn (1977). * considering the uncertaintyrelations” (48-49). A similar assertion is made later(118).
Bonotue
13. Compare withtheforce, lucidity, and succinctness of Kuhn’s Structure.
 sah
236 Chapter Fourteen

FS Sieaero
Dialogues or Paradigms? 297
arguments for the consistency of quantum theoretical descriptionas
ar- principle, classical physics is a limiting case of quantum physics—as
guments for its inevitability (chapter 9).
Yet it is not the details of Hanson’s argumentbut his conclusion we approach higher quantum numbers, quantum results mergeinto
that classical ones. The uncertainty principle, however, implies that state-
is the focus of my considerations: because a violation
of the uncertaint ments exist that are meaningful in the classical realm yet meaningless
relationsis a “logical impossibility,” quantum mechanics simply canner
be modified—it can only be totally overthrown: “The uncertainty prin- in the quantum domain (for example, the exact determination of a
ciple may of course be given up, but this would not be a particle’s position and momentum). Howcan then meaningless empiri-
reshuffling of
one or two elements at the top of the pile of micro-physical cal assertions transform into meaningful ones as we approach higher
knowledge: quantum. numbers?’
the wholestructure of that pile would collapse. ... One cannot ABEc
h This technical objection seems to threaten the whole edifice of quan-
a quantum-theoretic position andstill aspire for the day whenth
edif. tumphysics. Forit seemsto imply that either the correspondenceprin-
ficulties of the uncertainty relations will have been overcome. .
To ciple or the uncertainty principle must be given up. In Hanson's words,
hold a quantum-theoretic position just is to acceptthe relation
s as hi either “(1) quantum physics cannot embrace classical physics as a
avoidable” (1958, 149). As Kuhn’s paradigm forbids any dissent
from limiting case; or (2) quantum physics ought not to be regarded as per-
within, so Hanson's “conceptual pattern” forbids all but wholesale
change—bothnotionsare holistic in a strong sense and both discour manently restricted such that no analogue for the classical S [meaning-
- ful sentences] is constructable.”'” Hanson clearly rejected the second
age local criticism... . . “
What would be an example of such an overarchingalternative? option, which impliesthe incompleteness of quantum physics— the po-
Han- sition ofthecritics of quantum theory: “Alternative (2) has been chosen
son, with reservations, presented Bohm’s theory as a possible candi-
date.* Observations and experiments cannot decide as yet between by many eminent physicists and mathematicians: Einstein, Podolsky,
Bohm,de Broglie, and Einstein on one side, and Heisenberg, Born, and Rosen, de Broglie, Bohm”(1958, 153). Does it follow, then, that we must
Dirac on the other (1958, 174). The situation in quantum physicsis ieee. choose the first option—abandon Bohr’s correspondence principle?
lar to the incommensurable ways in which Kepler and Tychointer- No, argued Hanson,if we properly qualify the meaning of the corre-
spondenceprinciple using the notion of linguistic incommensurability.
preted the astronomical data of their time.
In another, more technical discussion, Hanson developed the no- The languages of quantum and classical physics, he explained, are —
tion of incommensurability as linguistic untranslatability. This “discontinuous”—there is no “ultimate logical connection between
devel- them.”'® Languages of “so different conceptual structures” cannot
opment, too, is aimed at defending the Copenhagen interpretation
against criticism. Some physicists and mathematicians, notably Her- simply mesh. When weincrease quantum numbers, quantum mechani-
mann Weyl, argued that the two central principles of the quantu cal sentences do not merge continuously into classical statements; in-
m stead, there are “formal analogies” between the expressions in the
theory —Heisenberg’s uncertainty principle and Bohr's correspondence
principle—contradict each other. According to the correspondence quantum andclassical realms. The fact that we use the same mathe-
matical symbols, or formulas, in the classical and quantum languages
obscuresthefactthat they are logically distinct (for example, weuse the
14. Hansonusedfor his discussion Schrédinger’s idea of
a wave packet, with the usual
Copenhagen excuse: “Though Schrédinger was wrong... there
is an advantage in adopt- 16. A similar objection applies when we express the same problem in termsof prob-
ing his exposition provisionally” (1958, 147). Hanson then asserted,
without any a
that matrix mechanicswill “issue in exactly equivale ability distributions. “Classical theory allows joint probabilities of accuracy (in deter-
nt results” (148). The excuse for ‘it
providing any proofis the usual-one of the “formidable” difficulty of mining pairs like time-energy and position-momentum), and allows them to increase
the formalism of
matrix mechanics. Yet the equivalence of matrix and wave simultaneously. In quantum theory this is illegitimate. But as quantum numbers get
mechanics was used by Han-
son, as by Géttingen-Copenhagen physicists, in a misleading way. The formal mathe- larger, the legitimacy of these joint probabilities seems to increase; the same perplexity
matical deduction of the uncertainty relations from the matrix arises” (Hanson 1958, 152).
formulation refers to the
statistical spread of measurement results. Therefore, it leaves open the question 17. Hanson mentioneda third possibility, which he immediately dismissed: “Classical
ofthe
exact determination of position and momentumfor an individual particle. physics itself should be regarded as restricted against the construction of S, justasis
15. Hanson mentioned doubts raised by Heisenberg, Oppenheimer, quantum physics” (1958, 153).
Dirac and Bethe
about the correctness of Bohm’s theory (1958, 174). He reminde 18. Hanson might have beeninspired by Bohr’s assertion that an absolute “cut” exists
d the reader that similar
arguments byde Broglie were “pulverized” byPauli; yet despite the hopeless betweenthe classical and quantum domains.
ness of the
“hiddenvariables” project, “Bohm remains undaunted”(173). 19. On the correspondence principle as “formal analogy,” see the excellent treatment
byDarrigol (1992a).
AUT + ques ae

298 Chapter Fourteen

Dialogues or Paradigms? 299
signs + and — in both numbertheory and valence theory, but thereis
of symmetrythat have been “the power and glory” of quantum theory
no logical identity between them). So understood, the correspondence
(1959, 10).
principle is not threatened by the consequencesof the uncertaintyfor-
Hanson’s defense of Copenhagen wastriggered by a draft of a favor-
mulas (1958, 49-157). able reviewof Bohm’s theory by Paul Feyerabend (1960). The existence
This linguistic untranslatability Hanson projected back into
history, of the Bohmianalternative convinced Feyerabend,as it would Bell after
to Kepler and Tycho andtheir supposed inability to communicate: “The
him, that the Copenhagen philosophy of indeterminism and subjec-
conceptualdifferences here reflect earlier examples: Kepler and
Tycho tivity was nota logical necessity (Heisenberg’s, Bohr’s, and Hanson’s
at dawn .. . , Beckman’s problem and Descartes’ problem,
Kepler's rhetoric notwithstanding) but a deliberate theoretical choice. The Boh-
law([s] as they were for him and then for Newton, .. . for Mach..
. and as mian alternative did indeed inspire Feyerabend's philosophy ofplural-
Hertz” (1958, 156). This statement reveals where the strange
idea of ism. It was this pluralism that Hanson rose to strangleat its very birth.
Kepler’s and Tycho’sinability to communicatecomes from. Yet Hanson
Feyerabend regarded Bohr’s interpretation not as a necessity but only
was not merely projecting the then current situation in physics ahistor
- as a metaphysical addition to the “bare” theory. Thus, according to
ically into the past. He was projecting a certain version —the
ortho- Hanson, Feyerabend concluded that we are free to inventalternatives
doxy— thatwas to serve as a “lens” through which he would interpr
et to Copenhagen. Hansonobjected strongly to this reasoning: “Let us
the whole history of physical thought. We see the strategy by which,
suppose that Feyerabend is.correct: would it followthat admitting this
intentionally or not, the stand of the orthodox, through supposedly dis-
implies that we are .. . free to invent and consider other metaphysical
interested philosophical discussion, is canonized into an overarching
interpretations? Not atall!” (1959, 5).
theory of the growth of knowledge. Hansonfirst ised quantum theory
Hanson’s conclusion was sharp: “Certainly no reinterpretation yet
to interpretthe past, to “describe” the essence of the “miscommunica-
suggested by philosopher or physicist presents a case for abandoning
tion” between Tycho and Kepler;.then he projected the reconstructed Bohr’s views.” Bohm was Hanson's direct target: ‘Thereislittle practi-
past ontothe situation in quantum physics: “It is the sense in which
cal warrantfor the alternative interpretations which have, since Bohm,
Tycho and Kepler do’not observe the same thing which must
be been receiving prominence” (1959, 1). Then Hanson judged even the
grasped if one is to understand the disagreements within microphys-
question of an alternative interpretation unwarranted, defending the
ics” (1958, 18). History itself, by this trick, comes down withall its
force “puzzle solvers” who unquestioningly follow the Copenhagenline:
on the side of the current orthodoxy. , “No one should think that because most quantum physicists are unper-
But is not my assumption that Hanson was defending the orthod
oxy turbed bythe tvpe of question brought to prominence by Bohm,that
too extreme? If my reading of Patterns of Discovery admittedly
involved therefore they are unreflective, resigned, . . . predicting machines”
some interpretive effort, Hanson's zeal to protect the orthodoxy
is (1959, 15).
crudely obvious in his paper “The Copenhagen Interpretation
” (Han- Mydiscussion of Hanson’s philosophyandits connection with Kuhn’s
son 1959), written in the year Patterns was published. Hansons
pelled ideas reveals why Kuhn wassuch a “reluctant revolutionary.”It sheds
out his intentions clearly: his paper “aims . . . to argue for orthod
ox light on what seemed a puzzling reaction by Kuhnto therelativistic
quantumtheory asit now stands” (1959, 1). In this paper Hanson
’s de- waythe tenets of Structure werelater interpreted. As for Hanson,sofor
fense of the irrefutability of the Copenhagen dogma reache
da high Kuhn incommensurability and the existence of different theoretical op-
pitch: “There is . . . no working alternative to the
Copenhagen Inter- tions did not imply relativism—there is only one preferred paradigm,
pretation. Ask your nearest synchrotron operator” (1959,
5). Follow- to the exclusion of all others. We can now better appreciate Kuhn’s ob-
ing Heisenberg faithfully, he attacked all those physicists
who dared jections to arguments that translated his notion of incommensurability
question the Copenhagen dogma. He joined Heisen
berg in discred iting into irrationalityof theory choice. We can also see, having analyzed the
“yesterday’s great men whohave not offered one scrap
of algebra to connection between Hanson and Heisenberg, the source ofthe strik-
back up their grandfatherly advice (Einstein, Schrédinger,
von Laue, ing similarity between Kuhn's “paradigms” and Heisenberg’s “closed
de Broglie, Jeffreys),” as well as those physicists who
offered simply theories.”
“pure mathematics” (Janossy) and “hidden variab
le” physicists who Unlike Hanson, Kuhn,as far as I know, never mounteda direct de-
offered a formalism that does not preserve the mathem
atical properties fense of the Copenhagen dogma. Yet revolutions are revolutions and
 300 Chapter Fourteen
Dialogues or Paradigms? 301
they demandtheir victims. Kuhn had
litt] i
feated. Asfor the alter
natives to Capenkaaen, Kuhn,likeH and far-reaching theory ofscientific development. Incommensurabil-
little doubt who deserved to disa ansen. had ity, [have argued, however, did not grow organically from Kuhn’shis-
ppear into the dustbin of Histon:
Those whorejected Newtonianism proclaimed torical scholarship—it was swiftly, perhaps hastily, superimposed on
innate forces would return science to that its reliance u S Kuhn’s emerging Structure, mostlikely as a result of an encounter with
the Dark Ages. Those wie 6 ‘
posed Lavoisier’s chemistry held that Hahson’s work.It is perhaps because of this abrupt introduction that
the rejection of chemical ’ fin
ciples’ in favorof laboratory elements was the rejec Kuhn struggled with the meaning of incommensurability throughout
chemical explanation. ...A simil tion of achieved his life: “Even before Structure appeared, I knew that my attempts to
ar. . feeling seems to underlie
opposition of Einstein, Bohm, and others, to the domi th. describe its central conception [fimcommensurability] were extremely
interpretation of quantum mechanics” nant probabilistic crude. Efforts to understand and refine it have been my primary and
(1970, 163). .
For’Kuhn, Bohm’s theory is not
a genuine alternative— it is an increasingly obsessive concern for thirty years” (Kuhn 1993, 315, my

ns beSN ll
table nuisance, if not a betrayal. irri- italics).7}
Feyerabend, who discussed Struc
with Kuhn (and whose “far-reaching and ture Having uncovered the ideological underpinnings of incommensu-
decisive” contributions to the
final draft of Structure Kuhn acknowledged rability, we should hardly be surprised that historians of science findit
—1970,xii), recalled: “I re-
member very well how Kuhncriticized increasingly difficult to build their narratives in Kuhnian terms. Past
Bohm for disturbing the uni-
formity of contemporary quantum theo scientific developments do not imitate the paradigmatic example of
ry” (1970, 206). Not surpris-
ingly, in his critique of Structure, Feyerabe quantum theory—not even, as Ihave argued throughout this book, the
nd mentioned a zoncentnal
similarity between the notion of a para development of quantum theoryitself. Rather than dogmatic commit-
digm andthesituation in quan:
tum theory: “Does he [Kuhn] want every mentto a rigid set of ideas, scientific creation—be it during the Coper-
subject to imitate the mono- nican, chemical, or quantum revolution—is often characterized by the
199):character of, say, the quantum
theory in the ’30s?” he asked (1970,
ingenious mingling andselective appropriation of ideas from different
Ironically, Bohm himself, wrote a “paradigms.” , BE sae
very favorable review of Structure
when it first appeared, and he harbored some As welearn from Robert Westman’s essay (1994), Copernicus him-
hopes (naive, we must
call them) that through Structure scientists self combined heliocentric and geocentric ideas, considering at first.a

ioe era
would “become aware of model very similar to Tycho Brahe’s. Tycho’s geometric system also was
the role that paradigms actually play in
thelife of scientific research the product—historical and conceptual—of an interface between two
in order that they shall be able more easil
y to realize the need for “incommensurable paradigms’—the Ptolemaic and the Copernican.
a change” (1964, 379). Years later, a
disillusioned Bohm had become
aware ofthe tight dogmatic implications of The pre-Structure Kuhn himself defined the Tychonic system as “an im-
Kuhnian paradigms, and he mediate by-product’ of Copernicus’s De Revolutionibus, despite the fact

i
€ I lange
g d I L. 1S Ini
) ti al enthu
siasm for Sti uctur é int
Oo 5 ob er crit icism
(B 0. hm. that “Brahe himself would have denied this” (1957, 205).
In fact, during the sixteenth century some Wittenberg astronomers

‘cmaakahoead Salaee Modis tucaSAth Ciara ASitAtaeek ben

Paradigms and the History of Science accommodated certain Copernican theses to the traditional geostatic
framework. Far from finding “communication” between Kepler and
The notion of incommensurable paradi ms Tycho impossible, astronomers incorporated Keplerian orbits into the
was, I h.
ceived by Hanson in the image of.the ect Tychonic system*—a development that the post-Structure Kuhn
uee of thequantercok
doxy. Through pretense, disguise, or unref
lective belief in disinterested
philosophical analysis, philosopherscane 21. During these years, Kuhn gaveup the notion of Gestalt conversion, yet retained his
asily objectify the interests of feeling that following a revolutionary change, a scientist experiences totally different
certain power groups, scientific
or political. Kuhn subsequently took world, “incommensurable”with the old one.
the notion of incommensurability as the
focal point ofa eartiprehenaive 22. “The use madeof Kepler’s elliptical hypothesis shows the samekind of selective
appropriation that can be found already in the assimilation of Copernicus’s planetary
20. Feyerabend
2 himseself
lf argued elo quently for incom models in the sixteenth century. Not everyone who accepted Keplerianellipses in the
i mensurabiliity between Ari
lian and Galilean physicsin Against Metho - seventeenth century also adopted the terrestrial motions, nor was anyone ‘converted’
d (1975); he claimed, however that ser
ae
of “incommensurable” changree are “rare pin by new planetary theory in the SSR [Structure] sense of conversion. The astrologer and
re’ ” in
ij the histor
i y of scienc
i e (Feyerabend 1981a,
French Royal Professor of Mathematics Jean-Baptise Morin (1583-1653), for example,
managed to accommodate Keplerian ellipses in a Tychonic frame. In this he followed a
 302 Chapter Fourteen
Dialogues or Paradigms? 303
perceived as a conceptual impossibility: “What is
regularly ignored . What Kuhn’s description nicely demonstrated washis initiation into
is the elementary fact thatelliptical [Keplerian] orbits
would have been the historian’s state of mind, gaining respect and empathyfor the in-
useles s if applied to any geocentric astronomical schem
e” (1971,
138 tegrity of past science, learning how to enter into the “shoes” of past
myitalics).
a thinkers, realizing that past knowledge was ingeniously and carefully
Turning to another revolution—that in chemistry—we
again find crafted. It was not an argument for inevitable holism. In fact, when
muchgreaterflexibility and selective appropriation
than the notion of describing howthe idea of a unique central earth was “interwoven”
paradigm changeentails. Lavoisier himself initially combi
nedthe phlo- (rather thanlater interlocked”) with fundamental concepts “within the
giston theory with his emerging ideas on the role of
air in combustion fabric of Aristotelian thought” (1957, 84) before Structure, Kuhn was
and calcination (work by Perrin, as discussed in Mausk
opfin press).3
Anoth er field of research—relativistic gravitational physi more careful: “Earth motion does not necessitate either the existence of
been explored byJohn Earman(1986, 1993). What
cs—has a vacuum. or theinfinity of the universe. But it is no accident that both
might look from text- these views wonacceptanceshortly after the victory of the Copernican
books like a monolithic paradigmis in facta plural
istic field, character- theory” (1957, 90).
ized bya diversity of opinions and approaches. ‘
Finally, I have argued throughoutthis book that in the ‘Kuhn’‘s argument in The Copernican Revolution hasall the irresistibility
quantumrevo- of a rational reconstruction. Yet, as my previous discussion indicates,
lution, mutual borrowing of techniques from compe
ting camps and paradigms are not impenetrable. Neither are they necessarily inco-
even of incommensurable” imagery characterize
d the creative efforts . ‘herent if modified piecemeal. The following example will suffice: Ac-
of Heisenberg, Schrédinger, Dirac, and Born. The emergenceo
f Heisen-_ ‘cording to Bohr, the acausality and contextuality (dependence on the
berg’s uncertainty principle was inspired by Schréd
inger’s wave im- experimental setup) of measurement results are intrinsically linked.
agery, which Heisenberg publicly denounced; Schrédinger’
s solutionof Bohr’s numerous thought.experiments supposedly showedthestrict
the radiation problem wasinspired by Heisenberg’s matri
x approach logical connection between the two. In Bohm’s version of quantum
which Schrédingercharacterized as “frightening and repuls
ive.” ‘ theory, the contextuality of measurementis retained yet the inevitability
As wehave noted, Heisenberg and Schrédinger were
closer on cer- of acausality is challenged—Bohm’s theory can be deterministic. There
tain issues than Heisenberg and Bohr (chapter 6); just as
Kepler was is nothing incoherent in Bohm’s framework—atleast no more so than
closer to the “anti-Copernican”Tychothanto. theCoperni
can Giordano in the Copenhagen one. Bohm’s version of quantum theory can,if
Bruno (Westman1994, 95).
one wishes, be constructed as a modification of the Bohrian paradigm,
Paradigms and Holism
rather than as a radical alternative to it. Where then does the notion of
“inevitable” holism come from?
The notion of holism underpins Kuhn’s incommensu Whendiscussing Heisenberg’s and Hanson's philosophy,I indicated
rability. Parad
igms
are holistic because “they require a numberofinterrelated the strong ideological drive behind the claim of holism for “closed theo-
changes of
theory to be madeat once.” Only at the price of incoherenc ries” and “paradigms”—it was designed to discourage critical ques-
e could these
changes be made piecemeal, one step at a
time. Kuhn illustrated
tioning of the orthodox view. I have also argued that Kuhn perhaps
his
pointby describinghisfirst-encounter with Aristotelian illegitimately translated his respect for the integrity of past knowledge
science which
at first seemed “dreadfully bad.” Only when Kuhn into the notion of a binding holism. Yet the assertion of the holism of
understood Aris-
totle on his own terms and realized how the Pieces paradigms might have another source—the holism implied by the
of Aristo telian
physics “fit together,” giving each other “mutu Duhem-Quinethesis.
al authority and sup-
port,” did he understand the nonarbitrariness As Robert Westman (1994) has noted, the overall image of science in
and holism of ja
lian physics (1987, 8). Kuhn's Copernican Revolution (1957) resembles some of Duhem’s theses,
and Kuhnhas acknowledged the “formative” role Quine’s “Two Dog-
strategy similar to those mid-sixteenth-century Witten mas” played in the development of his own philosophy of science
berg astronomers w ho accommo-
dated Copernicus’s equantless devices into a geostatic
reference frame” (Westman 1994, (1993, 313). The Duhem-Quinethesis asserts the intrinsic holism of
105-6).
/ 23. Tam gratefulil to
to S. S. M. M. Mauskopf£,
theoretical frameworks: an experiment does not contradict a single
wh shared with
pf, who i me hisi manuscript onthe histo-
riographyof Lavoisier’s studies, “The Chemical Revolution.” theoretical statement (which consequently can be saved from refuta-
° histo tion) but undermines the whole theoretical framework at once.
 304 Chapter Fourteen
Dialogues or Paradigms? 305
Yet this holism does not imply logical connection
s between different wayto grow,either in scienceorlife. Flexibility and open-mindedness
aspects of the theory. Thefact that a theoryis holist
ic in the sense that might prevent the accumulation of troubles altogether. Thesituation in
one rather than another aspect can be disproved does
not mean thatall my field—thehistory of science—can serve as an instructive example.
aspects standorfall together. Quite the contrary —the
fact that we can Muchofthe historiographyofthe pastthirty years has been informed
choose at will which one ofthe elementsis suspect
without undermin- ‘by thé Kuhnian notions of revolution and paradigm. Historians, using
ing the others meansthattheoretical frameworks do
not havethelogi- Kuhn’s concepts, increasingly got into trouble—first the notion of Ge.
calcoh esion impli ed by the notion of a holistic paradigm. Linda
sels‘s recent analysis of the implications of Bell’s theor Wes- stalt switches and the concept of conversions had to be abandoned,
em can be taken even by Kuhn.’ Yet the notions of revolution and paradigm were re-
as a nice illustration of this point. Wessels enumerated
a list of
the us- tained as adequate general descriptions of scientific change. Adherence
ual philosophical assumptions about systems in intera
ction and asked to these ideas has led historians to perform an astonishing varietyof ad
whichoneof these assumptions must be given up in
viewofBell-type hoc moves: Wehave a “quiet” revolution in chemistry. We have revo-
experiments (1989, 83, 88). It is enough to give up
onlyone of these lutions of Russian type and revolutions of Franco-British type (Gillies
deeply entrenchedideas, rather than all of them at
once. Which of the 1992a, 5): the Copernican revolution was “Russian,” while the Ein-
assumptions one in fact discards depends on thelocal, theore
tical, and steinian was “Franco-British.”** In mathematics, revolutions are only
sociopolitical circumstances.
of the Franco-British type (Gillies 1992a, 6). We have revolutions that
It is thislack of a binding holism that creates room forfruitful
dia- last no less than 300 years (the Newtonian), and others thatlast only (!)
logues betweenscientists with different approaches, andfor the
selec- 150 years (the probabilistic revolution; Kriiger, Gingerenzer, and Mor-
tive appropriation of ideas from opponents. From a dialogical
per- gan 1987). We also have somewhatsterile scholarly debates: should we
spective, the lack of paradigmatic holism is basic to the freed
om (not call certain developments revolutionary even if they do not fit the
arbitrariness!) of scientific theorizing.
Kuhnian framework (Gillies 1992b)?
Do wehaveto accumulate anomalies to such a degree before we can
Paradigms andCreativity
begin to question our concepts? Would we not be better off if we had,
from the beginning, been opento the adequacy of other types of marra-.
Kuhn’s notion of dogmatic commitment to a parad
igm was proposed tives? The idea that dogmais needed to prevent the wasting of effort is

ie 0% Aas, Lids Rtenb PESplueniad

not onlyas an adequate description ofscientific activit
y; it has a promi- built on ah erroneous presupposition—that the notion of a paradigm
nent normative dimension as well. Paradigmatic scienc
e is “science as is adequate. If different options are constantly open to negotiation,
it ought to be done” (Masterman 1970, 60).74 Only
singleminded, con- examination, selective appropriation (as I have argued is often the
centrated, consensual effort can lead to progress.
Without the rigid case), there is no duplication ofeffort or dispersionof creativeuerey:
commitment of its practitioners, science can easily slip
into wasteful Dispose of the notion of rigid paradigms, and the notion of “waste

a
anarchism—ofthe sort Feyerabend sometimes, tongue in
cheek, ad- disappears as well. In fact, opposition is vital for advance—not only aa
vocated. Kuhn never changed his position: it is the conser
vatism of competition or incentive, but as a reference point around which one’s
science,its rigid commitmentto paradigms, that produces,in
‘
his words, ownideasare elaborated, modified, and strengthened. -
“the greatest and mostoriginal bursts of creativity of
all human en- Of course, a certain amount of perseverance is necessary for the
deavors” (quoted in Horgan 1991, 15).
Dogmatic adherence to paradigmsensures “rapid success of any project. Yet perseverance does not imply dogmatic
scientific advance”
(Kuhn 1977, 232) because it alone allows the “loci 25. As Westman (1994) has pointedout, in Copernican Revolution most “conversion”
of trouble” (234) to
be restricted to those adamant points that refuse stories promised by Kuhn remained untold. ;
to surrender elegantly
to authoritarian paradigmatic methods. When such 26. In the Russian typeof revolution the old concepts and patterns of understanding
trouble spots, or are “irrevocably discarded.” In 1917 the tsar was executed, and the monarchyeliminated.
“anomalies,” accumulate, then a turning point—ther
ealization of the In theBritish and French revolutionsof the seventeenth and eighteenth centuries, respec-
need for a new paradigm—comes. Scientific revolu
tion is inaugurated. tively, the monarchies were overthrown andthe kings executed, but the monarchies were
An accumulation of troubles is not the only and perhap eventually restored. In the Franco-British type of revolution, the “old” persists but ex-
snot the best
periences “considerable loss of importance.” According to Gillies, the Copernican and
24. Kuhn endorsed Masterman’s dictum (Kuhn 1970, chemical revolutions were of the Russian type, while the Einsteinian was of the Franco-
postscript). British type (1992a,1, 5).
 at
canmansn egs

306 Chapter Fourteen

commitment. Similarly, mastery of one’s tools is necessary in any

CHAPTER 15
field—but does it mean that the use of scientific tools (mathematical,
experimental) is limited to puzzle solving? And whereas the impor- e7
tance of puzzle solving should not be underestimated, shouldit be glo-
tified? Scientific activity is not merely puzzle solving,it is also problem
Dialogical Philosophy and Historiography:
solving. Yet without alternatives, problems sometimes cannot even be
defined. Without Bohm’s challenge, we would not have Bell’s results A Tentative Outline
(Bell 1966; 1987, 160).
The creativity and longevity of what Kuhncalled “normal science” is —
S
due, I suggest, not to its dogmatism butratherto its lack ofit. The gesil- Dissonance,
ience of normalscienceis possible becauseit fits the Kuhnian notion of if you are interested,
leads to discovery.
normalscience poorly. In contrast, scholars have found that Kuhn’s no-
tion of rigid paradigms fits political groups and religious communities William Carlos Williams 1936, 176
especially well (see the papers in Gutting 1980).
Kuhn’s notionof the scientist as a puzzle solver seems especially apt To be means to communicate.
as a description of the experience of graduate and postgraduatestu-
dents in science—such as Kuhn himself was before he turned-to the Mikhail Bakhtin 1979, 312
history of science.” The experience of a mature creative scientist (nota
scientist in a “mature”science) is perhapsbetter characterized by David
Finkelstein: “Physics presently functions -with many of its practitioners
simultaneously seeking and doubting the existence of what they seek”
(1987, 291, myitalics). The uncertaintyofscience, its open-endedness, and the lively disagree-
This simultaneous seeking and doubting, when it occurs within a ment among its practitioners are sources of never-ending vitality. Sci-
web of stimulating, unrestricted dialogue,.reaps a rich scientific har- ence is a remarkable human achievementbecauseofits ability to turn
vest. The dialogical approach, in contrast to the Kuhnian structure both uncertainty and disagreement (supposedly disadvantages) into
of incommensurable paradigms and paradigmatic commitments, per- formidable strength. This ability to turn weakness into strength is due
ceives communication betweenfriends and foes alike as the precondi- to the basic responsivenessof scientific efforts—a responsivenessthat
tion ofall scientific creativity. Unlike the concept of a paradigm, which is institutionalized into an immense and amazinglyfruitful system of
by its nature excludes the “other,” the dialogical approach celebrates scientific communication. This responsiveness, combined with ever-
the existence of other minds as indispensable for scientific advance and growing experimental ingenuity and mathematical sophistication, un-
for intellectual self-identity. In science,asin all of our endeavors, giving derlies the unparalleled achievementofscience.
up openness, tolerance, individuality, and freedom gains us nothing. The dialogical analysis I articulated in part 1 of this book agrees re-
markably well with the view of science as seeking and doubting simul-
27. “But both my ownexperience in scientific research and my readingof thehistory of taneously, as doing one’s best today while being awarethatit all can be
science lead me to wonder whetherflexibility and open-mindedness have not been too overthrown tomorrow. When the freedom to doubt is suppressed or
exclusively emphasized as the characteristics for basic research” (Kuhn 1977, 226,
my circumscribed, the conditions for the emergence of dogmatic ideologies
italics).
prevail in scienceas in politicallife (part2).
Observingthat science is practiced in a communicative network, with
cultural, political, and institutional dimensions,is to state a platitude of
current studies of science.’ Noris there any dearth of excellent historical
studies of scientific dialogues and controversies. However, the use of

1. For the most recent example of the wide-ranging contextual studyof science, see
Galison (1997).
 cna
mina siihy 4anse

308 Chapter Fifteen

Dialogical Philosophy and Historiography 309
rigid terms for the description of science, even if reduced to
a local Dialogism as an epistemology of science is built on
“here” and “now,” does not permit an adequate expression of the notion of
fruitful the “other.” In the dialogical approach, the “social” is not
scientific freedom and constructive doubt. Such terms as “concep necessarily
tual shared or consensual, rather the social is that process or concep
schemes” and “frameworks,” when combined with the realizat t that
ion of by its nature assumesthe existence of the “other.” Dialogism does not
theoretical freedom (the underdetermination of theory by data, the
theory-ladenness of observation), lead to the notion of proliferation denythe sociological aspect of science, nor doesit merely supplement
of the basic cognitive substratum with “social factors.” The approach of
frameworks, to relativism and arbitrariness (Feyerabend 1975).
In contrast, the dialogical approach acknowledges the freedom to reducing the “cognitive” to the “social” is also alien to dialogism. In
theorize yet does not presuppose “conceptual frameworks” and thus the dialogical approach, the social and the cognitive are fused in the
~_ does notleadto their plurality. When science is practiced in communi- notion of an addressed thought andin the notion of a communicative
cative interaction, there is no serious reason to expect the existenc creative act.
e of Neither positivist nor post-Kuhnian conceptions of science have any
independent, let alone incommensurable, theoretical alternatives (chap-
ter 14). epistemological function for the “other.” Thus there is no place in the
While the idea thatscienceis practiced in a complex communicative current historiography of science for scientific individuality. The no-
tions of agreement and consensus, on either the global or the local
networkis a commonplace, the basic notion of the dialogical approach
developed in this book—the multidirectionality and addressivity of a scale, underlie most existing scholarship in studies of science and soci-
ety. While in pre-Kuhnian historiography agreement among scientists
single elementary thought—is far from being acknowledged, let alone
assumed,in existing accounts of the practice of science. was thought inevitable because most scientists followed the rational
method,” -post-Kuhnian scholarship examines the social mechanisms
leading to consensus. And while excellent historical studies of contro-
In Praise of Disagreement
versies and disagreements abound, they assume that the state of dis-
I propose to supplementcurrent approachesin the history and philoso- agreement is a temporary one—the disagreementis Sneenseduential
phyof science with the dynamic idiom of the “flux of addressivity.” I after the “experiments end” (Galison 1987), once “stormy waterssettle
(Collins 1992, 4), or when the controversy is “closed (Latour-1987).. .
proposeto treat the addressive response as the primary epistemologi-

ais
cal and social unit for the analysis of science. Thus the notion ofa sci- In the dialogical approach, the existence of the “other establishes the
entific thought presupposesthe existence of an interlocutor to whom preeminence of the concept of disagreement. This is not to say that
agreement does notexist in science, or is not valuable in the growth of

pit
the thought is addressed or by whose statements the thoughtis trig-
knowledge. But the explanatory power of the notion of agreementis
gered. Such “addressivity” is complex and multichanneled; scientists
respond to many colleagues and at the same time address other, limited. Excessive attention to consensus has become barrier to the
potential and actual interlocutors. This open-ended, multidirectional, construction of a dynamic theory of scientific practice. Too much reli-
constantly changing web is the source of the conceptual tensions and ance on notions of consensus hasresulted in a great loss to contempo-
ambiguity that fuel creative scientific effort. I adopt the term “dialo- rary historiography—the almosttotal extinction of studies of individ-
gism”for this dialogical approach to the formation ofscientific knowl- ual scientific creativity and of analyses of detailed theoretical labor. Too
edge(later in this chapter I will comment on the source ofthis term). much emphasis on agreement, combined with the philosophical thesis
Dialogism is a process epistemology, not a monological structural that theory is underdetermined by experimental data, has resulted in
theory of the growth of knowledge. The notion of addressivity is not the excesses of the sociology of knowledge and the flattening of the cog-
a static concept; it is both directional and temporal. It introduces the nitive to the social by social constructivists.>
locality, contextuality, and historicity of scientific practice in a funda- 3. The thesis of the underdetermination of theory by data is based on the realization
mental way. that a finite set of experimental results does not determine a unique scientific theory.
Many theories compatible with the same dataare possible. Philosophers who refuse to
2. WhatI call the multidirectional web of addressivity is in fact a time-vary drawrelativistic conclusions from this analysis employ the notion of epistemic “virtues”
ing net- (simplicity, consistency, generality) to single out a preferred theory. The founders of the
work, where by multidirectionality I mean the connectivityof the “I’’-node
to other nodes
representing different interlocutors. I minimize the use of the term sociology of scientific knowledgerightly point out that such epistemic virtuesfail as ex-
“network” becauseit
has some connotationsin the field of studies of science that I wish to avoid. planatory categories for the description of scientific practice (Bloor 1981). They argue in-
stead that the choice of scientific theories is determined by social categories such asin-
 310 Chapter Fifteen
Dialogical Philosophy and Historiography 317
The notion of uniform agreement, or consensus, is rooted in promi-
“belief.” Scientific solidarity occurs whenall participants arrive at the
nent sources in the philosophy, sociology, anthropology, and history same belief—either by the uncontaminated useof their reasoning
of science. In addition to Kuhn (1970), these sources include such im- fac-
ulties (“unforced agreement”) or by “brainwashing,” propaganda, or
portant names as Douglas (1975, 1986), Fleck (1979), Habermas (1972), “mob psychology.” * The notions of“belief” and “commitment” are no
James (1907), Peirce (1931-35), Rorty (1979, 1991), and Wittgenstein
less problematic than the notion of agreement. Scientists appear to be
(1953). Those thinkers who aimedto replace the bankrupt notion of too opportunistic, they often betray their “beliefs” and break their
truth as correspondence with reality employed instead the notion of “commitments,” deviating from codes of scientific methodology and
intersubjective agreement (Rorty 1991). They perceived such a consen- from collective norms. The notion of belief underlies the older, “inter-
sual, community-based vision of science as having clear humanistic nal’’ history of scientific ideas, and its mirror image—the strong pro-
advantages. In Rorty’s (1991) work, science provides models of “un- gram in the sociology of knowledge. In the strong program, “beliefs,”
forced agreement” and “humansolidarity.” Yet the notion of paradig-
which are socially, rather than cognitively, determined, are no less fun-
matic science is incompatible with the virtues of tolerance and open-
damental than in “internal,” purely intellectual, history.
mindedness. Byits very definition, a paradigm excludes diversity of
It is widely assumed in the current historiography of science that the
opinion and impugnsthe search for viable alternatives. ] have argued
onlv viable alternative to the solitary knower is the scientific collective,
that the notion of a paradigm legitimates the orthodox andsilences the held together by shared belief and mutualtrust. Another assumptionis
opposition (chapter14).It is only a totalitarian society that speaks with that such “trust” is mandatoryfor the formation of “sharedbelief,” and
one voice. * for the existence of the “social and cognitive order” (Shapin 1994). Yet
Noris the notion of agreement between mindsless problematic than another basic notionin the historiographyof science—boththepositiv-
that of agreement between “idea” and “reality.” Do we have any reli- ist and the sociological—is the notion of “commitment.” Those scien-
able epistemological tools to establish the existence of such agreement? tists who have strong “beliefs” are “committed”to their statements. For
The testimony of people whosay they “agree”is, of course, insufficient. how is one to be trusted if one does not believe in (if one is not com-
Bohr’s delightful expression “Weagree more than you think” exempli- mitted to) what oneis doing?”
fies this point.* The notionofintersubjective agreement seems to meas . The dynamic sociology of knowledge should not rely on static no-

Sts serbiabaoalae
deeply flawed as the cotrespondence theory oftruth: tions of “fixed belief” and “commitment.” We have to disassociate the
Therole of a scientific community in the Kuhnian and post-Kuhnian idea of devotion to one’s work from that of belief, perseverance from
historiography of scienceis to create and to implement agreement. Yet commitment, cooperation from consensus, trust from conformism. In
in this case, the scientific collective is epistemologically superfluous.® a multidirectional dialogical response onecan strike a creative balance
Moreover, something odd is going onin thescientific community all between trust and doubt, acceptance and skepticism, agreement and
the time. The morehistorians exploreactualscientific practice, the more disagreement. To reinstate the centrality of the notion of disagreement,
they discover that controversy is the orderofthe day.If the aim is agree- we have to strip away its negative connotations. Intellectual disagree-

dtex asd b abecds

ment, if the only way to progress is agreement, why doscientists spend ment need not be, andoftenit is not, quarrelsome or aggressive. While
so much time in long, sometimes bitter confrontations? Should not the
conflict among global “paradigms,” “conceptual schemes,” or “sys-
aims and the meansof scientific practice be in somewhat better har-
tems of thought” might imply threatening competition for survival,
mony? Rather than explaining away the disagreements as anomalies,
and the notion of pluralism of frameworks might suggest “wasted ef-
we need a theory of knowledge in which disagreement findsits proper,
fort” (Kuhn’s opinion, see chapter 14), the notion of local disagreement
prominent, and permanent epistemologicalplace.
Closely connected with the notion of consensus is the notion has none of these dark overtones. Local disagreement is fruitful and
of constructive, rather than threatening and wasteful.
terest. They also claim that they do notintend to eliminate the importan
ce of cognitive 6. In the social constructivist approach,the notion of “truth” is identical with the no-
aspects. The problem is not withtheir intentions but with the dearth of conceptua
ltools tion of “shared,” or “fixed” belief: “For historians, cultural anthropologists, and socio-
for integrating the social and the cognitive.
Baise

4. Bohr’s younger disciples in Copenhagenchose this expression logists of knowledge, the treatmentof truth as accepted belief counts as [a] maxim of
as the epigraph for method, and rightly so” (Shapin 1994, 4).
the Journalof Jocular Physics, in which they expressed their disagreem
ents with Bohr in a 7. “How could coordinated activity of any kind be possible if people could not rely
subtle, good-hearted fashion (Beller 1999),
5. I elaborate on this pointlater in this chapter, when discussing upon other’s undertaking? There would be no point to keeping engagements . . . with
the “lesser” scientists. people who could not be expected to honor their commitment” (Shapin 1994, 8).
 312 Chapter Fifteen
Dialogical Philosophy and Historiography 313
The notion of agreement permeates some of the most
exciting work The Philosophical and Historiographical Advantag
in the history of science. Shapin and Schaffer have provided a pioneer- es of Dialogism
ing accountof the origins of modern science as a social process aimed
Dialogism describes all novelty, both dramatic and relativ
at ending dangerous disputes (Shapin and Schaffer 1985; Shapin 1994), ely routine,
by the same mechanism of communicability. There
They describe how Boyle created distinct “technologies” for generating’ are no irrational
“Gestalt switches,” or jumps, in such a description. In
collectively agreed upon “matters of fact.” Yet what Boyle simulta this sense, dia-
ne- - logism is an unfolding description of the growth of knowledge, yet
ously aimed to establish were the rules for disagreeing in a gentleman- not
a linear one. Dialogism retains the idea of the progress of knowl
like fashion: “how to speak of things with freedom,” and yet edge
in the without reducingit to one-dimensional accumulation.It also retains the
manner“of persons with civility.” Finding things to agree upon (“mat-
fruitful notion of the continuity (the conservative aspect) of science. The
ters of fact”) was possibly a necessary condition for the conduct
ing of notion of addressive response is both social and cognitive. It is in prin-
scientific inquiry, yet nota sufficient one. The sufficient condition,I sug- ciple local and open ended. The notion of “best informed temporary
gest, wasan institutional way to handle disagreement, thus turning a presupposition” is more suitable to the description of dynamic scien-
vice into a virtue, a predicament into an asset. Henry Oldenburg,
the tific activity than the sociological notion of “fixed belief.”
secretary of the Royal Society, who transformed his network of cor- In the dialogical approach,the historical, philosophical, and socio-
respondenceinto a scientific periodical, Philosophical Transactions, real- logical merge into a unified viewpoint, rather than being independent
ized that in order to get ahead one had “to clash good men’s heads perspectives divorced from each other. The multidirectionality of dia-
together.” What becamedistinctive of modern science was not so much logical responsiveness accounts for the spectrum of possibilities, from
the scientific “method,” or scientific “norms,” but the strongly institu- the internal history of “pure” ideasto the strong versionsofsocial con-
‘tionalized communication system (scientific periodicals, scientific gath- structivism, obtainable as one-dimensional projections. The social de-
erings) to handle disagreement—a system that one cannotafford to “terminism in the writings of sociologists of knowledge is not an ade-
ignoreif one is to survive as a professional in a scientific community. quate description because it is confined to a single dimension. The
This communication system increased the pace of scientific discov- description of acausality in quantum mechanics as an expression of the
ery exponentially, not merely because discoveries were no longerlost, zeitgeist (Forman 1971) therefore provides only a limited’ perspective.

|{
but becauseit set the mechanism of communicative interactioninto full The notion of zeitgeist is itself a monological notion. The emergence
motion.§ : " of Heisenberg’s verdict on the essentially acausal nature of the quan-
According to Rorty, scientific institutions can serve as exemplary tum formalism was a complicated story, characterized by vacillations
models for the rest of the culture because they “give concreteness and and about-faces, not a one-dimensional adaptation (chapters 2, 3, 4, 5,
detail to the idea of ‘unforced agreement’” (1991, 39). Yetit is less un-

isl, i NA ealtt Tamas

8, and 9).
forced agreement than unforced response—be it agreement or dis- The dialogical approach incorporates the main message of Kuhnian
agreement—that is a crucial part of the scientific way of life. Those and post-Kuhnianhistoriography that an adequate philosophy ofsci-
thinkers who identified consensus as a necessary condition for social ence must take into accountthe lessons of detailed sociohistorical stud-
order spoke of “our colonization of each other’s mind”as the “price we ies. Kepler’s statement(an epigraphto this book) can serve as an inspi-
pay for thought” (Douglas 1975, xx). Yet there is no need for such brutal ration: “The roads by which menarriveat their insights . .. seem to me
“colonization”; it is uncoerced intellectual engagement and the right of almost as worthy of wonderas these matters in themselves” (quotedin
free responsethat are the very preconditions of scientific thought. Dis-
Koestler 1960, 59). Yet to understand how people arrive at their knowl-
agreementis a necessary part of such a response. Full agreementstops
edge is not simply fascinating —it is directly relevant to the epistemo-
iy

conversation. logical status of a knowledge claim. A successful scientific result is one

8. Studies of the scientific revolution (in the sixteenth to seventeenth
that is strongly embedded in the practice of the day. This embedding
centuries) center
on experimentation, mathematization, and the political underpinnings of the “new is achieved by responding simultaneously to as manyother scientific
sci-
ence.” These studies do notin general deal with the establishmentof scientific statements (read in papers, heard in conversations, and so forth) as
commu-
nication systems, and with their impact on the emergingscientific effort. Such a
discus- possible—drawing on many possible supports (confirmations), tak-
sion is absent from two recent comprehensivestudies of the scientific revolution ing into account manypossible objections (refutations). Both Carnap’s
(Shapin
1994; Cohen 1994).
idea of confirmation and Popper's idea of refutation are, again, one-
reste! 314 Chapter Fifteen
Dialogical Philosophy and Historiography 315
ed

dimensional idealizations, and both can be present at the same timein

fluctuation paper in Dirac 1927, though he mentioned his oral discus-
a dialogical response. sions with Heisenberg). Or one cancite papers for reasons of etiquette
Karl Popper’s theory of the growth of knowledge, in contrast to the
alone (as did those scientists who cited matrix mechanical papers while
Kuhnian theory, incorporates the uncertainty and open-endedness of
using the wave theoretical approach). Citations mightalso be, as Latour
scientific theorizing admirably. Yet Popper's identification of “testabil-
has demonstrated, about poweralliances. Yet citations are predomi-
ity” with “refutability” is too narrow, as manycritics did notfail to
nantly about the emergence ofideas, about the birth of novelty.
notice: scientists do other things besides trying to falsify their own
. The art historian Ernst Gombrich (1963) has quoted one of his col-
theoretical constructs. Moreover, from Popper’s perspective, scientific
. leagues as saying that“all pictures owe moreto otherpictures than they
theorizing is fundamentally irrational, for there is no organic connec-
_- do to.nature.” By analogy one could say thatscientific papers owe per-
tion between the method (refutation), and ‘the goal (truth; Newton-
- haps more to other scientific papers than to nature.This is not to dis-
Smith 1981, 44-70). a count experiments in the developmentof science but to emphasize that
Dialogism does not suffer from this weakness. From the dialogical
" the evaluation and incorporation of experimentalresults takes place in
perspective, the method (dialogical response) and the goal (strongest
an overall communicative flux of ideas. In such a communicative web,
possible support for scientific result, taking into account as manv po-
scientists have the freedom to take an apparent incompatibility be-
tential refutations and confirmations as possible). are closely linked.
tween a theory and observational facts not as a refutation of the theory
The dialogical approach I am suggestingis not that of Lakatos( 1970) or © - but rather as a reason to reinterpret the meaning of the observational
Elkana (1977), in which dialogues take place between global theoretical |
statements (theory-ladenness of observation). Such was Heisenberg’s
frameworks, such as “research programs”or “metaphysical systems.” reinterpretation of the electron’s path as a sequence of discrete points
According to dialogism, such global constructs conceal rather than re- -subject to.uncertainty relations rather than a continuous trajectory
veal the basic communicative nature ofscientific creativity. (chapter 4). This freedom of interpretation does not imply arbitrariness,
In the dialogical approach there is an organic, intimate connection quite the opposite: the theory-ladenness of observation provides the
between the philosophical and sociological accounts of scientific ac- leeway that allows a new result to relate to a larger numberof loci
tivity. The form ofa scientific article that includes references to other
in the dialogical web, thus entrenching it morefirmly in the practice of
papers appears to bean artificial conventionin most theories of the the day.
growth of knowledge. In contrast,from the dialogical point of view, The scientific community in the dialogical approach is not a gray
2
citations mirror the basic mechanism ofcreativity. Citations find a 3 army of indistinguishable puzzle solvers. Dialogism reevaluates the
prominentplace in Bruno Latour’s (1987) accountof scientific activity. i role of the “lesser” scientists in the developmentof scientific know]-
Latour’s analysis of citation as a means by which ideas survive is a | edge. In both evolutionary and revolutionary accounts lesser scientists
penetrating description of the mechanism of the dissemination and a
z
are, epistemologically, dispensable. In internal accounts of the history
a
consolidation of knowledge. Yet Latour’s analysis is more about the 3
4 of science, the evolution of ideas—"the great chain of being’— occurs
politics of domination than aboutscientific cognition. His approach —< exclusively in the mindsof great scientific heroes. In Kuhnian accounts,
haslittle to offer if one wants to understand an individual creative act. 2
lesser scientists obediently solve the puzzles posed by the great scien-
The conflation of the political and the cognitive might sometimes take
tists, founders of the new paradigms. Expandthelife span of the great
place—but it need not andoften does not. Dialogues need not be, even j scientists long enough, and they would accomplish everything them-
if they sometimesare, militant struggles for ascendancy.Basic dialogi- ; selves. The scientific community is, epistemologically, superfluous. In
cal responsiveness is not about the elimination of an opponent. Such
responsiveness can be bothcritical and sympathetic. the dialogical approach thelesser scientists are as indispensable as the
WhenI say thatcitations in a scientific paper mirror the basic
mecha-
1 great ones. In chapter 4, I demonstrated that someof the most impor-
nism ofscientific creativity, I do not mean to imply that the list
tant insights pertaining to the formulation of the uncertainty principle
of ref- i
by Heisenberg belongto scientists eliminated from the history of quan-
erences tells the story in a straightforward way. One can,
for reasons
fmsAeais din Udaint easewtacaans>

of self-aggrandizement, conceal one’s sources—a common

tum mechanics—Sentfleben and Campbell.
cause of From a dialogical perspective, the scientific community is composed
priority battles. One may be unable to include all sources
because of living, feeling individuals, rather than cognitive robots endowed
some are not yet published (as Dirac was unableto list Heisenberg’s
only with reasoning faculties. From the perspective of philosophical
 SeeEE
316 Chapter Fifteen
Dialogical Philosophy and Historiography 217
accounts, which formulate the criteria for the growth of knowledge in
process of discovery andthat of justification. If we do not formulate eur
pure epistemic terms (“degree of confirmations”— Carnap; “verisimili-
‘SES PS

thoughts in a waythat is addressed to the concerns of others, we cannot

tude”—Popper; “positive heuristic’—Lakatos), the intrusion of emo-
expect them to respond. Heisenberg’s uncertainty paper was permeated
tions into scientific activity is a regrettable,if not incomprehensible, ab-
with his responses to Schrédinger’s contention that the matrix (Heisen-
erration. While emotions have a prominentplace in the sociologically
berg’s) version of quantum mechanics was soabstract and nonintuitive
informed history of science, their existence remains epistemologically
that it would impede further progressin physics (chapters 3, 4, and 5).
unsatisfactory because they are derived from sociological notions. Rob-
The need to meet Schrédinger’s challenge, and to persuadethe scientific
ert Merton (1957) connects emotionalbattles for priority with scientifi
c community of the great value of the matrix approach (and so ensureits
norms (originality) and scientific rewards (fame). In the strong pro-
acceptance and dissemination), was both a motivation and a presup-
gram, emotions are derived from the existence of “interests.”
position of Heisenberg’s intellectual efforts. Thus discovery and justifi-
In dialogism, affect and cognition are closely linked. Thereis a natu- ‘
cation in the dialogical approach arein fact the same kind of processes,
emo

ral, organic connection between emotion and cognition in the very act }
permeated with responsiveness and addressivity. 1
;
of the local dialogical response. If our intellectual activity is in prin- hi
This is not to saythat there is no distinction between the process of
ciple responsive to other humanbeings, then cognition and emotionare discovery and the resulting achievement. The elegant formulation of
often fused. The ideas we encounter are not “disembodied”; they are
matrix theory as an algebraic structure is different (a categorydiffer-
* colored bythe voice ofthe other. These voices are marginal or authori- ence) from the contrived path that led to its discovery. Yet this structure
tative, remote or immediate, threatening or sympathetic. This is the rea-
wasjustified, not by some general method comparing theory and ex-
son that even the most abstractscientific notions can arouse fierce emo-
periment, but rather by the same communicative process of using ma-
tions (chapters 2 and 4). Henri Poincaré observed that most creative
trix theory in further research. While thehistory of its birth is largely
mathematical ideas are “those which,directly or indirectly, affect most eliminated from the reported scientific result, a trained eye can some-
profoundly our emotionalsensibility” (quoted in Vernon 1970, 77-78).
times discern the traces of the processof discovery even in abstract

SikPE teal Se ENSeenaee

This emotionalsensibility can be connected with personal ambition, as mathematical formulas. Thus the elements of the position matrix are a
revealed in the confrontations between Heisenberg, Schrédinger, and reminder of the reinterpretation procedure that led- Heisenberg to the
Bohr (chapters 2, 4, and 6): emotions in scientific activity can be con- matrix formulation of quantum mechanics (chapter2).?
nected with great aesthetic satisfaction, as in Dirac’s case (Kragh 1982;
Darrigol 1992a), with meaningful existential issues, as for Bohr (chap- Theory as Practice: Between Tools and Metaphors
ter 12), or with religious and mystical issues, as for Pauli (Laurikainen
1988). While most recent historiography of science is hostile to indi-
Recently much workin the history of science has focused on the idea of
vidual psychology (concentrating instead on shared ideas, values, and
scientific knowledge as practice rather than as static atemporal repre-
norms), the dialogical approach centers on the creativity of individual
sentation. The new approach explores the multiplicity and heteroge-
scientists. Such individual psychology:should be based not on “pur-

dure
neity of scientific doing, describing the different aspects of scientific
ported eternal verities of the mind,” which historians, with good rea-
activity—the conceptual, the social, and the material—as fragmented
son, reject (Daston 1995, 4); rather it should study the individual’s com-
and local. A prominent philosophical advocate of this approachis Ar-
plex set of responses. Dialogues,be theysilent or vocal, are often fueled
thur Fine (1986, and especially the new postscript in 1996b). In addition
-and colored by individual emotions.
to Fine’s work, this trend owes much to Ian Hacking’s influential Rep-
Connected with the emotion-cognition issue is the philosophical
resenting and Intervening (1983). Hacking’s book has inspired increased
distinction between the “context of discovery” and the “context of jus-
attention to the notion of experiment, and to a thorough reevaluation
tification.” The process of discovery can be contaminated with emo-
of the complex role and natureof scientific experimentation. Hacking’s
tions, while the process of justification supposedly consists of objective,
rational evaluation (Reichenbach 1959; Popper 1959). This distinction is
increasinglychallenged byhistorians, philosophers, and sociologists of 9. Knowingthe heuristics of discovery canalso illuminate the physical understanding
of a theory. This is the reason Pauli declined Born’s offer to collaborate on the elaboration
science (Nersessian 1995; Nickles 1995; Knorr-Cetina 1981). From the of Heisenberg’s results into the matrix formulation: “You are only going to spoil Heisen-
dialogical point of view there is no essential distinction between the berg’s physical ideas by your futile mathematics” (quoted in van der Waerden1967, 37).
 318 Chapter Fifteen
Dialogical Philosophy and Historiography 319
workhas been followed by such important booksas Shapin and Schaf- this aspect in any way. The dynamic, performative nature of
fer (1985), Franklin (1986), Galison (1987, 1997), Buchwald (1994, 1995), scientific
theorizing is common both to the instrumental and dialogical aspects
and Pickering (1995). of science.
The dialogical approachis well suited to emphasize the dynamic as- When there was agreement in the community of quantum physi-
pect of scientific theorizing, and to minimize the static elements of the cists, it was more often about the efficiency of tools than about meta-
representational idiom in the description of the growth of knowledge. physical “paradigmatic” issues. Einstein did not doubt the unprece-
Theory, no less than experimentation, is practice: scientist
s engage in . dented utility of quantum mechanicaltools, only their philosophical, or
theorizing, writers of textbooks and philosophers describe “theorie
s.” rather ideological, interpretation by the orthodox. Einstein challenged
This is not to say that scientists do not use representations theCopenhagen philosophy by skillfully using the mathematical tools
and visual
imagery—graphs, photographs, computer images, diagrams, and vi- » -of the quantum formalism (chapter 7). Einstein especially disliked
sual models are indispensable scientific tools. Yet the meaning of such
-Bohr’s metaphor of complementarity, which he considered arbitrary
representations, like the meaning of observational terms, is embedded
and removed from scientific theorizing (chapter 12). Dyson similarly
f in the dialogical context. The idea of the de Broglie wave packet (super--
. places the revolutionary aspect of science in the construction of new .
posed waves with different wavelengths) as a representation of a micro-
tools rather than in philosophical ideas or metaphors. One can build
scopic particle was employed extensively in quantum theory and its
many different models, metaphors, and conceptual worlds with the
interpretation. Yet Schrédinger, Heisenberg; and Bohr used the idea of
same tools—such models and metaphors can be dispensable, while -
a wave packet each in his own way and each for-his own ends. Each the tools are not. The contrast between the solidity of tools and the
ascribed to the concept a different theoretical function and a different
naman ge 2

fluidity of metaphorical structures represents a fascinating tension in

degree ofreality, and each employed the wave packet imagery tocriti- ' _ the history of theoretical physics. The tools are used in everyday prac-
cize the use of this notion by the opponent(chapters4,6, and-11). ‘tice, while revolutionary philosophical ideas and metaphors are em-
The notion of theorizing naturally introduces a temporal elementinto 3
2 ployed primarily for dissemination and consumption (chapters 8, 9, 11,
PANES

~ historicaldescription; 1° in the dialogical approach theorizing inherently =

12, and 13).

ei ite NNSsRie,
takes time. In chapter 2, I described the emergence of Born’s probabilis-
Kuhn regarded a revolutionary change as metaphorical redescrip-
tic interpretation as a process “in flux,” during which creation, delib-
tion: “The juxtaposed items are exhibited to a previously uninitiated
eration, and reevaluation (all of which take time)occurred. In chapter4,
by someone who can already recognizetheir similarity, and who urges’
I described how Heisenberg’s uncertainty principle emerged from a
_ that audience to learn to do the same” (1987, 20). Bohr’s complemen-
coalescence of insights, arrived at in different dialogues. Heisenberg’s tarity was just such a metaphorical redescription. Yet Bohr did notfind
preference for acausality was shaped as a response to Schrédinger’s the existing similarities—he imaginatively created them (chapter 12).
competitive, causal version of quantum mechanics, to Campbell's dis-

ed Radlabaabs Suibe
Not surprisingly some of the most important textbooks on quantum
cussion of time and chance in quantum theory, to Sentfleben’s delibera- mechanics omit the complementarity principle (Dirac 1930; Tomonaga
tions about causality, and to Jordan’s doubts aboutthe essentially sta-
Bae 1962; Feynman, Leighton, and Sands 1969). The fact that quantum me-
ie tistical character of quantum theory. Causality issues also’surfaced
in | chanics is such a potent theory and yet has no uniformly accepted in-
Heisenberg’s dialogues with Pauli and’Dirac. Parts of these dialogu
es terpretation demonstrates that a coherent metaphysicsis not necessary

i eS, Lp,
occurred concurrently, while others occurred atdifferent times. We
see to the advancementof theoretical science.” That philosophical agree-
here not the unfolding of a single timeless argument, but the gradual emeeetesnceminalmeneame AA OLAS SAAN AAA INN erth WE alliaS

coalescence and confrontation of insights that, over

ment is dispensable in scientific theorizing is clear from Schweber’s
time, reinforced, (1994) masterful studyof the history of quantum electrodynamics. The
illuminated, and qualified each other. The notion
of practice invokes
the notion of tools (mathematical, experimental);
and while dialogism, 11. Though the application of tools is neither self-evidentnorself-explanatory. Out-
as developed here, does not provide adequate epistem
ological under- side a dialogical web, formulas are mute. While the impossibility of the simultaneous
pinnings for the instrumental aspectofscience, it does not
contradict determination of position and momentum was implied in the basic commutation formula
Pq ~ 4p = (h/27i)1 from the beginning, it was only mathematically derived later by Hei-
10. Thereal-time understanding of practice is, as several senberg, during his struggles with the interpretation of quantum mechanics.
authors argue, a major histo- 12. This conclusionis akin to Duhem’sattitude, andto that of Duhem’s most prominent
riographical challenge today (Pickering 1995; Stump 1996).
. follower today, van Fraassen.
 320 ChapterFifteen
Dialogical Philosophy and Historio
graphy 321
outstanding theoretical physicists who founded quant
um electrody- say), still at each moment, in a contingent way, he or
namics—Dyson, Feynman, Schwinger, and Tomonaga—h she is exposed
ad less mo- only to a part of the existing world ofresults. Moreover, the scientist’s
tivation to build easily digestible, visualizable models
and explanations responses— the threads—are not predetermined.Still
for wider consumption thantheir prominentpredecess more freedom
ors, whoerected comes into play whenthescientist weavesthese responses into a theory.
the nonrelativistic quantum mechanics.
. Yet this imaginative, skillful weaving is not arbitrary—thecreative sci-
To say that metaphorsplay a limited role in theoretical physic
s does entist attempts to use as manythreads as he or she has at hand each
not imply that the search for metaphorsis superfluousor illegi
timate. moment and aimsto constructastight, and as strong, and as structured
Metaphors can be powerful heuristics. In addition, the need
to find a tapestry as possible. In this way the achieved result can be both per-
some broader meaningin scientific results, fluid as.
they are, the need to suasive to others and beautiful. The processof creation (weaving) and
reconcile one’s intellectual activity with one’s spiritual life is
one many the result itself (tapestry) are both imaginative and well grounded.
scientists feel. It is this need that accounts for the wide appeal
otf Bohr’s Thus, in a dialogical process, scientific theorizing is both free and
complementarity (chapter 12).3 Yet complementarity,
I have argued, nonarbitrary. It is this nonarbitrariness that underlies a scientist's feel-
is not merely an inspiring metaphor; it is also a potent
ideological ing that his or her theories are “touching”reality itself, that the best
weapon. Bohrandhis followers imposed rigid strictures on the
imagi- results are “true” and “objective.” Dialogism respects the scientist's
nation of physicists, claiming that complementarity is not simply one
feeling that his best work is as well grounded (“objective”) as humanly.
possible wayto interpret quantum physics but the only possible way
(chapter 9). possible. Thus, while rejecting the philosophical notion of theories as
yp eS “mirrors” of reality (the correspondence theory of truth), dialogism is
compatible with scientists’ intuitions of a high degree of truth and ob-
Truth and Beauty jectivity of their constructs. By concentrating on creative individuality,
dialogism does not consider science subjective.
Seemingly irreconcilable tensions characterize the pronou
ncements Notall aspects of scientific theorizing are equally well grounded—
physicists make aboutscientific theorizing. On the one hand,they talk
about nonarbitrariness, well-groundedness, and the near
some are more arbitrary than others. While Heisenberg’s uncertainty
PTR OPR erenpence ae

inescapability formula and Schrédinger’s wave equation are indispensable for the
of scientific results. On the other hand,they describe the feeling
of free- working quantum physicist, he or she can create without the Copen-
dom, imagination, beauty, and even “poetry” that underlies
creative hagen interpretation. Thus the Copenhagen interpretationis dispens-
scientific effort. Those philosophers who soughttheories of individ A
ual “8
able—many physicists engaged in successful scientific theorizing are
rationality and a bindingscientific method did not leave epistem Pu
ologi- 4
opposed to it, while manyothers subscribingto it nevertheless never
cal room for free individual creativity. Sociologists and social histori
ans use it in their research. It is reasonable to consider such dispensable
of science dispensed with the idea of an overarching scientific
method, parts of science more “subjective” and arbitrary than the indispensable
yet their accounts often imply too much cognitive arbitrariness.
Dialogism can incorporateboth the freedom and the nonarbitrarine parts. It is illuminating to analyze these aspects of science in termsof
ss the rhetorical strategies of persuasion (part2). _
of scientific work. The editors of The Flightfrom Science and Reason,
sum- While closely dealing with the details of scientific ideas, dialogism
marizing the paper by Hershbach (1996), compare scienceto “a tapestr

wie bes
y can be considered a social epistemology ofscience. It is social because,
of variable design, constantly under construction, with seemin
gly un- in its basic notions of addressivity and responsiveness, it presupposes
related threads combining in unexpected and beautiful
ways” (Gross, the existence of other minds (other people) as a necessary condition
Levitt, and Lewis 1996, 9). In dialogism, the threads
are the creative of scientific process. While concentrating on the content of ideas, the
scientist’s dialogical responses, directed or triggered by
the multiplicity notion of an overarching scientific method that guides an individual
of existing results reported by his or her peers. And while
- a working scientist is foreign to it. In this sense, dialogism is removed from the
scientist must be as informed as possible (in contrast
to creative artists, inspirations of logical positivists, followers of Popper, and those philos-
ae rere armen

ophers of science today who would resuscitate, in modified form, the

13. I distinguish between imaginative and vagueextensions
of scientific results by met- idea of a self-sufficient individual rationality. The notion of a Divine
aphors into wider domains(such as Bohr’s complementarit
y principle) and well-defined Architect, or a Divine Creator, into whose wisdom and intentions a
structural analogies used as heuristics (Bohr’s correspondence
principle) single scientific genius can penetrate (Newton, Einstein) is foreign to
PRESNparee
 322 ChapterFifteen
Dialogical Philosophy and Historiography 323
the dialogical approach (though dialogism recognizes such imagery
as twentieth-century physics as composed of different theoretical, experi-
powerful, motivational, and heuristic).
mental, and instrumental subcultures, each having a “life of their own.” |
Dialogism as social epistemology differs fundamentally from social
This collection of distinct subculturesis held together by “trading along
constructivism. Dialogism does not hold that social

eee
determinants, in their borders’”—such trading is nothingelse but local communicability, -
the sense of the usual sociological variables, are the primary determi
- which does not require overarching agreements and thus avoids the
nants of scientific ideas. Unlike social constructivism, dialogism does
problems caused by the use of homogeneousglobal frameworks (such
not conflate beliefs and facts—dialogism shies away from the
notion as relativism). -
of belief altogether. Dialogism rejects the social construc
tivist notion of Galison, following Peirce, offers a “cable” metaphorforthe descrip-
the consensustheory of truth: dialogism rejects the notion of consensus
tion of scientific process and scientific community —thedistinct, par-
as mandatory and the notionof truth as explanatory.
tially independent fibers (subcultures) are laminated and intertwined,
Dialogism byits nature paysclose attention to the specifics of the lo-
cal situation. For that reason, dialogism does not reject the contributions and the whole structure gains strength by their interaction and mutual
support. The cable metaphor proves most fruitful for Galison’s descrip-
heaps

achieved in other approaches,including social constructivism. Whether

‘tion of twentieth-century physics. ‘
sociological or political factors are predominant (eugenics, Lysenk
o While there is much similarity between Galison’s approach and the
affair), mildly relevant, or virtually nonexistent (Dirac’s theorizi
ng in dialogical point of view I am articulating (an emphasis onlocal inter-
quantumtheory)is to be determinedbythe specifics ofthe local theoriz-
_ activity, a rejection of global “blocks” of knowledge,the articulation of
ing. Dialogism accepts (and, in fact, requires) the important realization
strength following from multiplicity rather than homogeneity), there
of social scholars of science that scienceis practiced in rich variety of
are also crucial differences between the two. From the dialogical point
social, political, cultural, national, and institutional settings.
of view, a single strand of the cable (a subculture)is still too homoge-
’ Arthur Fine has recently formulated an agendafor a new,,adequate
neous; a moreapt metaphoris that of (quantum) entanglement—the
philosophyofscience:
ideas of different scientists are so “entangled”thatit is often difficult
1. Bracket truth as explanatory concept. to talk about the separate individuality of such ideas. Galison intro-
2. Recognize the openness of science at every level, especially the per- duces the cable metaphorin order to explain the cohesion of science:if.
vasive activities of choice and judgment. one decomposes global conceptual schemesinto local fragments, what
:
holds scientific practice together? From the dialogical point of view,
3. Concentrate on local practices without any presupposition as such a problem does not arise: the cohesive nature of the scientific en-
to
how they fit together globally, or even as to whether they do fit to-
bein

terprise, be it fragmented as it may, is assured bythe dialogical connec-

gether.
tivity of a single creative act. .
4. Rememberthat science is a humanactivity, so thatits understand- The dialogical approach developedin this book bears some similar-
ing involves frameworks and modalities for social action. ity to the ideas of Mikhail Bakhtin, the Russian literary theorist and
5. Finally, on the basis of all the above, try to understandthe phe- philosopher, whose writings I encountered after most of this book was
nomenaof opinion formation and dissolution in science in all its par- written. The Bakhtinian approachis based on the concept ofdialogue
ticularity. (1996a, 249) ~ , and the indispensability of the notion of the “other” in human existence
Wecan conclude that dialogism meets Fine’s challenge.“
and thinking. Every thought, according to Bakhtin, is conditioned by
es pieMaebeARIEL

The fruitfulness of the notion of “disunity” in science is powerf the actual and potential thought of the other. Weare alive to the extent
ully that we are responsive (and responsible), as well as addressive to the
illustrated in Peter Galison’s (1997) recent book. Galison
describes other.
_ 14. With regardto item 1: bracketing the notion
of truth does not mean thatscie
In his classic Problems of Dostoyevsky Poetics, published in 1929, Bakh-
ntific
theories are arbitrary and do not adequately reflect nature;
rather it means that the gen- tin (1994) singled out Dostoyevsky as the discoverer of a newliterary
eral abstract notion of “truth” has proved bankrupt for the
description of science. With
regard to item 4: Dialogism doesnot, at this stage, offer more
generalsocial modalities, 15. While Bakhtin has a large following in literary theory, history, and other human
yetit is basically “friendly” to such an undertaking.
The notion ofbasic
responsiveness, whether explicit or not, probably underli addressivity and
és any theory of social action.
j and social sciences, his ideas have had hardly any impactin the history and philosophy
; of science.
ii
mere
 324 Chapter Fifteen
Dialogical Phiiosophy and Historiography 325
genre—the polyphonic novel. Dostoyevsky’s novels and short stories
is no difference between scientific andliterary texts. The difference is in
display the open-endedness andirreducible plurality of human exis- fact profound. While bothscientists andartists respond to the works of
tence. Dostoyevsky’s heroesare not clearly defined characters— their
others and address their audiences, working scientists cannot afford to
truth, never a completedone, emerges in confrontation with,
and antici- be mute(if they are to survive as professionals) in the dialogical flux of
pation of, the other’s point of view. Everyidea, those of the characters ideasin their field. In particular and perhaps most important, theorists
and those of the author,is a rejoinder in unfinished and unfinalizable
must respond to (and incorporate) the experimental findings reported
dialogues carried out simultaneously. The author does not have an
ob- by their colleagues.
jectifying privileged gaze: she or he can unfold the open-ended life of The creative artist is not bound by what might becalled a “dialogical |
characters byletting them.speak, yet cannotdefine or explain them. imperative.” Ifa working scientist withdraws from intense engagement
Bakhtin contrasts such polyphonic thinking with the monologism of with the recent workof others, his or her own work may be not merely
natural science. According to Bakhtin, in science there is no genuine
superseded but even declared irrelevant. Thus, in addition to experi-
placefor the other: the otherin scienceis the one whoerrs. In contrast, in
mentation and mathematization, arts and sciences are distinguished by
art and spiritual life, two voices are the minimum for thought and exis- the difference in the extent and necessity of dialogical communicability.
tence. The Bakhtinianscholar Holquist (1990) coined the term “dialog- Dialogism as developed here does not (and possibly cannot) ade-
ism” to describe Bakhtin’s approach. I adoptedthis term for a dialogical quately coverall aspects of scientific practice—dexterity and ingenuity
epistemologyand historiographyof science, despite Bakhtin’s opinion in using scientific tools, whether experimental or mathematical, andin-
that the dialogical approach doesnot apply to the natural sciences. Con-
strument making may be outsideits application. Nor doesthe dialogi-
tra Bakhtin, I argue that the “other”is fundamental in science no less,
cal approach permit a complete reconstruction ofallfacets of creativity
and arguably even more, than in other human endeavors. in scientific theorizing— intellectual feats of imagination,risk taking,
Dialogism, as developedin this book, differs substantially from Bakh- and serendipity remain an enigma.
tin’s point of view. Whileit is a powerfultool for the analysis of scien-
Nevertheless, the awareness of dialogical communicability allows a
tific texts (chapters 5, 6, and 7), the dialogical approach to science goes deeper and more detailed understandingof the (rational) workingsof
beyond textual analysis. For Bakhtin, “the author” was a potent con- a scientific mind. Dialogism is not a general “theory of the growth of
struct for the analysis of poetics and not a creative historical figure. knowledge”; ratherit is an approach, an attitude, a realization that can -
Dialogism in science attempts to accurately reconstruct the process of
inform andI hope prove valuable in the study ofscience.
discoveryby creative scientists. As I have mentioned, many scientists are aware of the communica-
Nevertheless, the dialogical approach, with its awareness of the ad-
tive nature of their own endeavors (chapters 1 and 5). Such awareness
dressivity of our thinking, whether in science or in art, allows us to
pervades Schrédinger’s reflections on his scientific activity in his No-
establish a meaningful bridge between the sciences and the humani- bel lecture: “In my scientific work (and also in mylife), I have never
ties. There is notable similarity between my analysis of Heisenberg’s
followed one main line, one program defining adirection for a long
uncertainty paper (chapter 5) and Bakhtin’s analysis of Dostoyevsky’s time. ... My work is never entirely independent,since if ] am to have
novels. As a novel by Dostoyevskyis shot through with rejoinders and an interest ina question, others mustalso have one. My word is seldom
anticipatory responsesto the other, so discourse in a scientific paper
the first, but often the second, and maybeinspired by a desire to con-
carries an awareness ofaninterlocutor, through the author's anticipa- se
tradict or to correct, but the consequent extension may turn out to be
tion of how the other might approach the problem. As in The Brothers
more important than the correction” (quoted in Darrigol 1992b, 238).
Karamazov each of Ivan’s speechesis a dialogue not only with Alyosha
Schrédinger’s words eloquently capture the importance of disagree-
and himself but with otherpoints of view that are addressed simulta-
ment, addressivity, and open-endedness in scientific theorizing. Or, in
neously, so Heisenberg’s paper is permeated with addressivity to the
Bakhtin’s phrase, “there is neither the first word nor the last word”
“conservative Lords,” to Pauli, Dirac, Bohr, Jordan, and others.
As a (1979, 378)osesscs
novel by Dostoyevskyis a plurality that does not achieve resolution,
so
a scientific paper is an open text whose meaning will undergo change
in future dialogues.
This similarity should notbe takentoofar. It does not meanthatthere
 REEER EN'GES
Ey

rec, ae
Abbreviations

- AHQP Archive for the History of Quantum Physics, American

Philosophical Society, Philadelphia, assembled and edited
by T. S. Kuhn,J. Heilbron, P. Forman, and Lini Allen.
Sourcesfor History of Quantum Physics: An Inventory and
Report, Philadelphia: American Philosophical Society .
APHK Bohr, Atomic Physics and Human Knowledge (1958)
ae
ATDN Bohr, Atomic Theory and the Description of Nature (1934)
BCW Niels Bohr Collected Works, followed by volume number.
Essays Bohr, Essays 1958-1962 on Atomic Physics and Human Knowl-
edge (1963)
PE Hermann et al., Wolfgang Pauli, Scientific Correspondence with
Bohr, Einstein, Heisenberg (1979)

Aharonoy, Y., and D. Bohm. 1961. “Time in the Quantum Theory and

aeitins i OR tN es SP aadta ie al iBi

the Uncertainty Relation for Time and Energy.” Physical Review 122:
1649-58. Reprinted in Wheeler and Zurek 1983.
Aris, R., H. T. Davis, and R. Stuewer, eds. 1983. Springsof Scientific Cre-
ativity. Minneapolis: University of Minnesota Press.
Aronowitz, S. 1988. Science as Power: Discourse and Ideology in Modern
Society. Minneapolis: University of Minnesota Press.
Asquith, P. D., and R. N. Giere, eds. 1981. Proceedings of the 1980 Biennial
Meeting of the Philosophy of Science Association. Vol. 2. East Lansing,
Mich.: Philosophy of Science Association.
Badash,L., T. O. Hirshfelder, and M.P. Broida, eds. 1980. Reminiscences
ofLos Alamos, 1943—45. Dordrecht: Reidel.
 es oi voilil
328 References
References 329
Bakhtin, M. 1979. Estetica Slovesnove Tvorchestva.
Moscow: Isskusstvo. . 1997b. “Criticism and Revolutions.” Science in Context 10:(

is ae), bila
. 1994. Problems of Dostoyevsky’ Poetics. Edited and transla
Se ERSweeRERTE

ted by 13-37. 1):

C. Emerson. Minneapolis: University of Minnesota Press.[ ee PER

Ree
First pub- . 1998. “The Sokal Hoax: At Whom Are We Laughis
lished in 1929.]
Ballentine, L. 1970. “The Statistical Interpretation of Quant Today, September, 29-34. or
um Mechan- . 1999. “Jocular Commemorations: The CopenhagenSpirit”
ics.” Reviews ofModern Physics 42:358-81.
Barnes, M., and D. Edge, eds. 1982. Science in Context ris, forthcoming. (an
: Readings in the . Forthcoming. “Kant’s Impact on Einstein Thought.”InHoward : oes
Sociology of Science. Cambridge, Mass.: MIT Press.
and Stachel, forthcoming. ..
Basso, K. H., and H.A, Selby, eds. 1976. Meaning in Anthrop
ology. Al- Beller, M., R. Cohen, and J. Renn, eds. 1993. Einstein in Context..Cam-»
buquerque: University of New Mexico Press.
Batens, D., and J. P. van Bendegem,eds. 1988. Theory and Experi bridge: Cambridge University Press. serine f
ment: Re- Beller, M., and A. Fine. 1994. “Bohr’s Response to EPR.” In Faye and
cent Insights and New Perspectives on Their Relation. Dordre
cht: Reidel. Folse 1994, 1-31.
Bell, J. S. 1964. “On the Einstein-Podolsky-Rosen
Paradox.” Physics 1: Beltrametti, E., and B. C. van Fraassen,eds. 1981. Current Issues in Quan-
195-200. tum Logic. New York: Plenum. 7 .
. 1966. “On the Problem of Hidden Variables in Quantum Me-
Bancbleoshern, Y. 1990. “Equivalent Descriptions.” British Journalfor the .
chanics.” Reviews ofModern Physics 38:447-52. af / ne Bea en
Philosophy of Science 41:261—79.
- 1987. Speakable and Unspeakable in Quantum Mechanics. Cam-
‘ 1992 arene with Realism: Schrédinger’s Case.” In Bitbol
he:
bridge: Cambridge University Press.
Gr: and Darrigol 1992, 25-40. Py

callie 3:8 06 BS aanidGoedel tease ibaa akMinesaces 1lA

Beller, M. 1983. “Matrix Theory before Schrédinger: Philosophy, Prob-
Bernhard, CG. E. Crawford, and Per Sérbom, eds. 1982. Selences Tech-

Se
lems, Consequences.”Isis 74:469—91.
nology and Society in the Time of Alfred Nobel. Oxford: Pergamon:
. 1985. “Pascual Jordan’s Influence on the Discovery of
Heisen-
berg’s Indeterminacy Principle.” Archive for the History of Exact
Beyler, R. H. 1996:Targeting the Organism: The Scientific and Cultural
Sci- Context of Pascual Jordan’s Biology, 1932-47.” ISIS 87: 248-73.
ences 33 (4): 337-49. History of Science
Biagioli, M. 1990. “Galileo’s System -of Patroriage.
- 1988. “Experimental Accuracy, Operationalism, and the Limits ot
of Knowledge, 1925 to 1935.”Science in Context 2:147-62. 28:1-62.
Bitbol, M., ed. 1995. The Interpretation of Quantum Mechanics. Wood-
- 1990. “Born’s Probabilistic Interpretation: A Case Study of a he - .
bridge, Conn.: Ox Bow.
‘Concepts in Flux.’”” Studies in the History and Philosophy ofScience 21:
1096. Schrédinger’s Philosophy of Quantum Mechanics. Dordrecht:
563-88.
Kluwer. a ;
. 1992a. “The Birth of Bohr’s Complementarity: The Context and
Bitbol, M., and O. Darrigol, eds. 1992. Erwin Schrodinger: Philosophy and
the Dialogues.” Studies in History and Philosophy of Science 23:147
—80. the Birth of Quantum Mechanics. Paris: Editions Frontiéres.
. 1992b. “Schrédinger’s Dialogue with Gottingen-Copenhagen
Black, M. 1962. Models and Metaphors. Ithaca, N.Y.: Cornell University
Physicists: ‘Quantum Jumps’ and ‘Realism.’” In Bitbol and Darrig _
ol Press.
1992, 277-306.
- 1993. “Einstein and Bohr’s Rhetoric of Complementarity.” Sci- Blaedel, N. 1988. Harmony and Unity: The Life of Niels Bohr. Madison:
Wisconsin Science Tech Publishers. .
ence in Context 6:241—55.
Bloch, F. 1976. “Heisenberg and the Early Days of Quantum Mechan-
me

- 1996a. “Bohm and ‘Inevitability’ of Acausality.” In Cushin

g, ics.” Physics Today 29 (12): 23-27. a
Fine, and Goldstein 1996, 211~30.
- 1996b. “The Conceptual and the Anecdotal History of Quan- Bloor, D. 1981, “The Strengths of the Strong Programme.” Philosophy of
ashenae

tum Mechanics.” Foundationsof Physics 26:545—57. the Social Sciences 11:199—213.

Blum, W., H.-P. Diirr, and H. Rechenberg,eds. 1984. Werner Heisenberg:
- 1997a. “Against the Stream: Schrédinger’s Interpretation
of Collected Works. Munich: Piper.
Quantum Physics.” Studies in History and Philosophy ofScien
ce 28B (3): Bohm, D. 1951. Quantum Theory. New York: Prentice-Hall. Reprint, New
421-32.
York: Dover, 1989.
 esis
330 References
References 331
. 1952. “A Suggested Interpretation of the Quantum Theory in
Terms of ‘Hidden Variables,’ J and IL.” Physical Review sion in ATDN, 52-91. German version in Naturwissenschaften
85:166-93. (1928): 245—57.] Page references are to Wheeler and Zurek.
- 1957. Causality and Chance in Modern Physics. London: Rout-
. 1929a. “The Atomic Theory and the Fundamental
ledge.
Underlying the Description of Nature.” Lecture for theS al
. 1964. “The Structure of Scientific Revolutions. By T. Kuhn.”
Meeting of Natural Scientists. Translated in ATDN, 102-1
Philosophical Quarterly 14:377-79.

>
. 1987. “Hidden Variables and the Implicate Order.” In
nally published in Danish in Fysik Tidsskrift, 1929. Facsimile oftr
Hiley.and lation in BCW, 6:236-53.]
Peat 1987, 33-46. . . 1929b. “The Quantum of Action and the Description of Nature
- 1996. On Dialogue. Edited by LeeNichol. London: Routledge.

>
(in German). Naturwissenschaften 17:483—86. Translated in AED

aha
Bohm,D., and B.J. Hiley. 1993. The Undivided Universe: Ari Ontological:
92-119. [Facsimile in BCW, 6:208-17.] Page references are to ATDN.
Interpretation of Quantum Mechanics. London: Routledge.
. 1930. “Philosophical Aspects of Atomic Theory.” Address to the

DineORE a SEdle
Bohm, D., and F. D.Peat. 1987. Science, Order and Creativity. New. York:
Royal Society of Edinburgh, 26 May. Unpublished manuscript. Bohr

ayia
Bantam. _ & & / Manuscript Collection, AHQP. In Sanders 1987, 129-46. [BCW has
Bohr, N. 1913. “On the Constitution of Atoms and Molecules.”
Philo- onlyfacsimile of an abstract from Nature 125:958.]
sophical Magazine 26:1-25, 476-502, 857-75.
—. 1931a. “Harald Hoffding.” Speech at the meeting of the Society
. 1918. “On'the Quantum Theory of Line Spectra”: Part I, “On:
of the Sciences, 11 December. Bohr Manuscript Collection, AHOP.
the General Theory,” Part Il, “On the Hydrogen Spectrum.” Det In Sanders 1987, 173-80. [Originally published in Danish in Oversigt
Konglelige Danske Videnskabernes Selskab, Matematisk-Fysiske Meddelser
over det Kongelige Danske Videnskabernes Selskabs Virksomhed, 1931] ,
4 (1): 1-36, 36-100. . 1931b. “Space-Time Description and Conservation Principles.
- 1924. “Onthe Application of theQuantum Theory to Atomic Notes for unpublished paper, 16 November 1930~—20 February 1931.
Structure, Part I.” Proceedings of the Cambridge Philosophical Society,
Bohr Manuscript Collection, AHOP. In Sanders 1987, 147- 60.
suppl., 1-42. . 1932. “H. Heffding’s View on Physics and Psychology.” Notes
. 1925. “Uber die Wirkung von Atomen Bei Stéssen.” Zeitschri
ft from lecture at the tenth International Psychology Congress, August.
fir Physik 34:142-57. Translated in BCW, 5:194-206.
Bohr Manuscript Collection, AHOP.In Sanders 1987, 195 ~203. Ai
- 1926. “Atomic Theory and Mechanics.” Nature 116 :845—52. . 1933. “Light and Life.” Address at the opening meeting o s e
- 1927a. “Fundamental Problems of the Quantum Theory.” Un-
International Congress on Light Therapy, Copenhagen, August. c
published manuscript, 13 September. Bohr Manuscript Collection,
ture 131:421-23, 457-59. Revised version in APHK, 3-12. [Danis
AHQP. In BCW, 6:75-80.
version in Naturens Verden 17:49—54. oe— in Naturwissen-
- 1927b. “Philosophical Foundations of the Quantum Theory.” :245—50.] Page references are to ;
Unpublished manuscript. Bohr Manuscript Collection, AHOP. and the Description of Nature. Cambridge:
In Seet ae te
BCW, 6:69-71.
Cambridge University Press. [Reprinted as The Philosophical Writings
. 1927c. “The Quantum Postulate and the Recent Develo
pment of Niels Bohr, Vol. 1, Woodbridge, Conn.: Ox Bow, 1961, 1987.]
of Atomic Theory.” In Atti del Congresso Internazionale dei
Fisici, 11— . 1935a. “Can Quantum-Mechanical Description of Physical Re-
20 Settembre 1927, 565-88. Bologna: Nicola Zanichelli. .
Reprinted in ality Be Considered Complete?” Physical Review 48 2G ~102.
BCW, 6:113-—36. Page references are to BCW.
- 1927d. “The Quantum Postulate and the Recent Develo . 1935b. “Space and Time in Nuclear Physics. Notes from lec-
of Atomic Theory.” Unpublished manuscript, 12-13
pment ture at the Society for the Advancementof Physical Knowledge, Co-
Bohr Manuscript Collection, AHQP. In BCW,
October 1927. penhagen, 21 March. Bohr Manuscript Collection, AHQP.In Sanders
6:89-—98. [Reprinted in
Sanders 1987, 75-89.]
— es” Analysis and Synthesis in Science.” International Ency-
- 1928. “The Quantum Postulate and the Recent Develop
mentof I nified Science 1 (1): 28.
Atomic Theory.” Nature 121 (suppl.): 580—90. Reprinted
and Zurek 1983, 87-126. [Facsimile in BCW,
in Wheeler Te Oe and onde Physics.” Address delivered at the
6:148-—58. Revised ver- Physical and Biological Congress in Memoryof Luigi Galvani, Bolo-
 332 References
References 333
gna, October. In APHK, 13-22. [Originall
y bli i j . 1957c. “Position and Terminology in Atomic Physics
the Galvani Congress, Bologna]. i .” Notes,
pabiishedin Vroceedings # 11 March 1957. Bohr Manuscript Collection, AHOP. In Sander
- 1937c. “Caus
Ca alityJ and Complementarity. s 1987,
y.”” Philos
Phi ophy of § i 665-67.
. en n [Originally published in German
in Erken
ntnis éwe - 1958a. Atomic Physics and Human Knowledge. New York: Wiley.
ngush e
e version repri
Pp nted in Sanders rs 1987 241-5
~ 2.] Page references: ‘ [Reprinted as The Philosophical Writings of Niels Bohr, Vol. 2, Essays
- 1937d. Hitchcock lectures at 1932-1957 on Atomic Physics and Human Knowledge, Woodbridge,
the University of California, Berke Conn.: Ox Bow, 1987.]
ley, 8-24 Marc -
h. Unpublished manuscript. Raymond
tion, American Institute of Physics. Birge Collec- :4 . 1958b. “Atoms and Human Knowledge.” Address at convoca: --
In Sanders 1987, 253-355 tion, Roosevelt University, 4 February. Bohr. Manuscript Collection,
- 1938a. “The Epistemological Problems
ofAtomic Theory.” 2 Ma AHQP. In Sanders 1987, 707-18. ;
1938. Bohr Manuscript Collection, AHQP. In Sanders1 ;
—: 1938b. “Natural Philosophy and 987, 377-80,» - 1958c. “Quantum Physics and Philosophy: Causality and Com-
Human Cultures.” Address plementarity.” In Essays, 1-7. [Originally published in Klibansky
deliv ered at the International Congress
of Anthropological and Eth- 1958. Reprinted in Sanders 1987, 679-87.] fe , 3
nological Sciences, Kronborg Castle, Elsinore,
August. In APHK, 23- . 1958d. “Rudjer Boskovic.” Talk at the Rudjer Boskovic Celebra-
31. Pnginally published in Comptes rendu a4
es du congrés international de tions, Zagreb. Bohr Manuscript Collection, AHQP. In Sanders 1987,
science, anthropologie et ethnologie, Copenhagen. 4
English version i 719-34. ne = &
ture 143 (1939) : 268-72.] ia
—— 1939. “The Causality Problem in Atom
os versione Naw . 1958e. “The Unity of Knowledge.” John Franklin Carlson lec-
ic Physics.” In New Theo- ture, Iowa State University, January. Bohr Manuscript Collection,
ries in Physics, 11-45. Paris: International
Institute of Intellectual Co- AHOP.In Sanders 1987, 689-705. , -
operation.
ea e Reprinted in Sanders 1987, 383—
422. Page references are . 1960a. “The Connection between the Sciences.” Addressto the

2
.- 1948. “On the Notions of Causa Internatiénal Congress of Pharmaceutical Sciences, Copenhagen, 7

ssNinMsGE te
lity and Complementarity
fae .” Di-2 August. Translated in Essays, 17-22. [Reprinted in Sanders 1987,
re on
alecti e
a Be A ee

ca 2:312 -19. Repr

p inted in Sanders 1987, 445~—
445-55. Page references 757-66.) ee 2 re -
1 “Discussion with Einstein on
. 1949. . 1960b. “Quantum Physics and Biology.” In Symposia of the Soci-
Epi
pistemological
i Problems ety for Experimental Biology. Vol. 14, Models and Analogues in Biology.
Ee
in Atomic Physics.” In Schilpp 1970, 201-41. Repri i Reprinted in Sanders 1987, 781-90. ,

tara ees 2G, VG bens Redpiceal ua Le AER

E and Zurek 1983, 9-49. ve Paiied in Wheeler . 1960c. “The Unity of Human Knowledge.” Address delivered.
- 1950. Summary of the Gifford lectures,
F ti
25 ga
Janua ry,1950. Bohr Manuscript
pt Colle
7 27 D ecember 1949~
ctition, AHQP. In Sanders
Collec
at a congress arranged by La Fondation Européennedela Culture,
Copenhagen, October. In Essays, 8-16. [Reprinted in Sanders 1987,
- :1955a. “Atoms and Human Knowledg 767-79. bee
[Also i In APHi
e.” . 1961. “The Genesis of Quantum Mechanics.” In Essays, 74-78.. -
in Daedalus 87 (1958): 164-74. ®°
ere
=
- 1955b. “Unity of Knowledge.” In APHK [Originally publisned in German as “Die Entstehung der Quanten-
, 67- igi mechanik,” in Bopp 1961.] oo
published in Leary 1955.] ‘ OEena . 1962. “Light and Life Revisited.” Address delivered at the in-
» ohes “Mathematics and Natural Philo
sophy.” Scientific Monthly auguration of the Institute for Genetics, Cologne, June. Unfinished
evatie eprinted in Sanders BE, 551-61. Page refer
ences are to manuscriptin Essays, 23-29.
- 1957a. Draft . 1963. Essays 1958-1962 on Atomic Physics and Human Knowledge.
s for a postscript to “Discussion
Epistemological Problems in Atomic with
with Einst e
Einstei NewYork: Wiley. [Reprinted as The Philosophical Writings of Niels
Physics [Bohr 1949] ” August. Bohr, Vol. 3, Woodbridge, Conn.: Ox Bow, 1987.]
Bohr Manu script Collection, AHOP. In Sande
rs 1987, 668-76. . 1972, 1984, 1985. Niels Bohr: Collected Works. Vols. 1, 5, and 6.
- 1957b. “The Philosophical Lessons
of Atomic Physics.” Karl
reamlaine Atom * cohenaoe do sass

Taylor Comptonlectures, Massachusett Amsterdam:North Holland. :

s Institute of Technology, 5— Bohr, N., H. Kramers, and J. Slater. 1924. “The Quantum Theoryof
26 November. Niels Bohr Holdings,
Niels
Institute of Physics. In Sanders 1987, 593-6 Bohr Library, American Radiation.” Philosophical Magazine 47 :785-—822. Reprinted in van der
48. Waerden 1967, 159-76. Page references are to van der Waerden.
 Ne UNAalsPANT SsaRRanRSascoe Sol
334 References
References 335
craresR soe?

Bohr, N., and L. Rosenfeld. 1933. “Zur Frage der Messbarkeit der Elec- . 1955b. “Statistical Interpretation of Quantum Mechanics.”
tromagnetischen Feldgrossen.” Det Konglelige Danske Videnskabernes
ne eee ae ei

Nobel Prize lecture. Science 122 (3172): 675-79. [Reprinted in Born

Selskab, Matematisk-Fysiske Meddelser 12. [English translation in Cohen 1956.
and Stachel 1979. Reprinted in Wheeler and Zurek 1983, 479-534] — Physics in My Generation. London: Pergamon.
= ae sreeears

Bopp, F,, ed. 1961. Werner Heisenberg und die Physik unserer Zeit. Braun- . 1961. “Erwin Schrédinger.” Physics B1 17:85-87.

|
schweig: Friedrich Vieweg und Sohn. . 1962. Physics and Politics. Edinburgh: Oliver and Boyd.
Born, M. 1924. “Uber Quantenmechanik.” Zeitschrift fiir Physik 26:379— . 1963. Ausgewahlte Abhandlungen. Gottingen: Vanderhoek and
96. Translated in van der Waerden 1967, 181-98. Page references are - Ruprecht.
to van der Waerden. _ . 1964. Natural Philosophy of Cause and Change. New York: Dover.
. 1926a. Problems of Atomic Dynamics. Cambridge, Mass.: MIT . 1969. Atomic Physics. London:Blackie.
Press. ; , ed. 1971. The Born-Einstein Letters, New York: Walker.
. 1926b. “Quantenmechanik der Stossvorgange.” Zeitschrift fiir Born, M., and W. Biem. 1968. “Dualism in Quantum Theory.” Physics
Physik 38:803—27. Translated as “Quantum Mechanics of Collision Today 21:51—56.
Processes,” in Ludwig 1968, 206-25. Page references are to Ludwig. Born, M., W. Heisenberg, and P. Jordan. 1926. “Zur OQuantenmechanik
- 1926c. “Zur Quantenmechanik der Stossvorgange.” Zeitschrift Il.” Zeitschriftfir Physik 35:557-615. Translated as’”“On Quantum Me-
fur Physik 37:863~67. Translated in Wheeler-and Zurek 1983, 92-555".

i iANEMT aba
Re RTSERRaNEa ID eS

chanics th’ in van der Waerden 1967, 321-85. Page references are to
Page references are to Wheeler and Zurek. . : van der Waerden. .
. 1927a. “Das Adiabatenprinzip in der Quantenmechanik.” Zeit- Born, M., and P. Jordan. 1925a. “Zur Quantenmechanik.” Zeitschrift
schriftfiir Physik 40:167-92. “oy ® fiir Physik 34:858—88. Translated in van der Waerden 1967, 277-306.
. 1927b. “Physical Aspects of Quantum Mechanics.” Nature 119: Page references are to van der Waerden.
354-57. Reprinted in Born 1956, 6-13. Page references are to Born . 1925b. “Zur Quantentheorie aperiodischer Vorgange.” Zeit-
1956. ' schrift fiir Physik 33:479-505.
. 1928. “On the Meaning of Physical Theories.” Nachrichten der . 1930. Elementare Quantenmechanik. Berlin: Springer.
8g SFE ees SORRY ENN:

Gesellschaft der Wissenschaften.zu Géttingen. Translated in Born 1956, Born, M., and N. Wiener. 1926. “Formulierung der Quantengesetze der
17-36. Page references are to Born 1956. - Periodisher und Nichtperiodischer Vorgange.” Zeitschrift fiir Physik -
—. 1936. “Some Philosophical Aspects of Modern Physics.” Pro- 36:177-87.
ceedings of the Royal Society of Edinburgh 57, pt. 1: 1-18. Reprinted in Bothe, W., and H. Geiger. 1925a. “Experimentelles zur Theorie von
Born 1956, 37-54. Page references are to Born 1956. Bohr, Kramers undSlater.” Naturwissenschaften 13 :440—4}.
- 1950. “Physics and Metaphysics.” Memoirs and Proceedings of . 1925b. “Uber das Wesen des Comptoneffekts: Ein experimen-
the Manchester Literary and Philosophical Society 91. Reprinted in Born teller Beitrag zur Theorie der Strahlung.” Zeitschrift fiir Physik 32:
1956, 93-108. Page references are to Born 1956. 639-63.
——. 1951. “Postscript to the Restless Universe.” Reprinted in Born
qe

Bridgman,P. W. 1927. The Logic of Modern Physics. London: Macmillan.

1956, 225-32. Page references are to Born 1956: . 1930. “Permanent Elementsin the Flux of Present-Day Physics.”
- 1953a. “The Conceptual Situation in Physics and the Prospects Science 71:21.
of Its Future Development.” Guthrie lecture, March. Proceedings ofthe Bub,J. 1992. “Quantum Mechanics withoutthe Projection Postulate.”
Physical Society A 66:501-13. Reprinted in Born 1956, 123-39. Page Foundations of Physics 22:737—-54.
referencesare to Born 1956. . 1997. Interpreting the Quantum World. Cambridge: Cambridge
- 1953b. “The Interpretation of Quantum Mechanics.” British University Press.
Journal for the Philosophy of Science 4. Reprinted in Born 1956,
140-50. Buchwald,J. Z. 1994. The Creation of Scientific Effects: Heinrich Hertz and
Page references are to Born 1956.
Electric Waves. Chicago: University of Chicago Press.
- 1953c. “Physical Reality.” Philosophical Quarterly, 139-149.
Re- , ed. 1995. Scientific Practice. Chicago: University of Chicago Press.
printed in Born 1956, 151-164. Page references are to Born
1956. Bukharin, N. I, ed. 1971. Science at the Cross Roads, new ed. London:
- 1955a. “Is Classical Mechanics in Fact Deterministic?” Physica-
Cass.
lische Blatter 11 (9): 49-54. {Reprinted in Born 1956.]
Butterfield, H. 1931. The Whig Interpretation of History. London:Bell.
aleneremaasatt NY
 sitab
336 References
:4 References 337
Campbell, N. R. 1921. “Atomic Structure.” Nature 107 (2684):17
0. i - 1994b. Quantum Mechanics, Historical Contingency and the Copen-
. 1926. “Time and Chance.”” London, Edinburgh and Dublin Philo-
hagen Hegemony. Chicago: University of Chicago Press.
sophical Magazine and Journal of Science, 7th ser., 1:1106-17.
i eae
. 1996. “The Causal Quantum Theory Program.” In¢
. 1927. “Philosophical Foundations of Quantum Theory.” Nature
Fine, and Goldstein 1996, 1-19. ae —
119:779, ~ j
4
Cartwright, N. 1983. How the Laws of Physics Lie. Oxford: Clarendo
Cushing,J. T., A. Fine, and S. Goldstein, eds. 1996. Bohmian Mechanics
n. and Quantum Theory: An Appraisal. Dordrecht: Kluwer.
- 1990. “Philosophical Problems of Quantum Theory: The Re- sae
Cushing,J. T., and E. McMullin, eds. 1989. Philosophical Consequen
sponse of the American Physicists.” In Kriiger, Gingeren
zer, and Quantum Theory: Reflections on Bells Theorem. Notre Dame, Ind:
Morgan 1987, 417-37. :
versity of Notre DamePress. e. i ES Cedar
Casimir, H. B. G. 1967. “Recollections from the Years 1929-1931.” In
Darrigol, O. 1986. The Origin of Quantized Matter. Waves.” Historical
Rozental 1967, 109-14.
Studies in the Physical Sciences 16, pt. 2: 197—253. “i ”
- 1983. Haphazard Reality: Half a Century of Science. New York:
- 1992a. From c-Numbers to q-Numbers: The Classical Analogy in the
Harper and Row.
History of Quantum Theory. Berkeley and Los Angeles: University of
Cassidy, D.C. 1992. Uncertainty: TheLife and Science of Werner Heisenberg.
California Press.
New York: Freeman.
. 1992b. “Schrédinger’s Statistical Physics and Some. Related
Chevalley, C. 1988. “Physical Reality and Closed Theories in Werner Themes.” In Bitbol and Darrigol 1992, 237-76. a.
Heisenberg’s Early Papers.” In Batens and van Bendegem 1988,
Daston,L. 1995. “The Moral Economyof Science.” Osiris 10:3-24.
159—76.
Daumer, M. 1996. “Scattering Theory from a Bohmian Perspective.” In
- 1994. “Niels Bohr’s Words and the Atlantis of Kantianism.” In
Cushing, Fine, and Goldstein 1996, 87-97. a
Faye and Folse 1994, 33-53.
Davidson, D. 1984. Inquiries into Truth and Interpretation. Oxford: Clar-
Cohen, H.F. 1994. The Scientific Revolution: A Historiographical Inquiry.
endon. .
Chicago: University of Chicago Press. de Broglie, L. 1923. “Ondes et quanta.” Comptes Rendus 177:507-10.
Cohen, R. S., and M. Wartofsky, eds. 1968. Proceedings of the Boston Col-
De Witt, B.,and N. Graham, eds. 1973. The Many Worlds Interpretationof
loguium for the Philosophy of Science 1966/1968. Boston Studies in the
Quantum Mechanics. Princeton, N.J.: Princeton University Press.
PhilosophyofScience, vol. 5. Dordrecht: Reidel. Dieks, D. 1994. “Modal Interpretation of Quantum Mechanics, Mea-
Collingwood, R. G. 1939. An Autobiography. Oxford: Oxford University
surements, and Macroscopic Behaviour.” PhysicalReview A 49:2290-
Press.
300. vs
Collins, H. 1992. Changing Order: Replication and Induction in Scientific
Diner, S., D. Fargue, G. Lochak, and F. Selleri, eds. 1984. The Wave Par-
Practice. Chicago: University of Chicago Press.
i Colodny,R. G., ed. 1972. Paradoxes and Paradigms: The Philosophical Chal-
ticle Dualism. Dordrecht: Reidel..
F Dirac, P. A. M. 1925. “Fundamental Equations of Quantum Mechanics.”
Ky: lenge of the Quantum Domain. University of Pittsburgh Series in the
bia Proceedings of the Royal Society of London’A 109:642-53. 00
od
Philosophyof Science, vol. 5. Pittsburgh: University of Pittsburgh
es . 1926. “Quantum Mechanics and a Preliminary Investigation of
ey
ie Press.
the Hydrogen Atom.” Proceedings of the Royal Society of London A 110:
SPEERmek

, ed, 1986. From Quarks to Quasars. University of Pittsburgh Series

561-79.
in the Philosophy of Science, vol. 7. Pittsburgh: University ofPitts-
. 1927. “The Physical Interpretation of Quantum Dynamics.” Pro-
burgh Press.
ceedings of the Royal Society of London A 113:621-41.
Compton, A. H. 1923. “A Quantum Theory of the Scattering of X-rays
. 1929. “Quantum Mechanics of Many Electron Systems.” Pro-
by Light Elements.” Physical Review 21:483—502.
ceedings of the Royal Society of London A 123:714-33.
- 1927. In Nobel Lectures in Physics 1965, 174-90. .
RN

. 1930. The Principles of Quantum Mechanics. Oxford: Clarendon.

cep

Crombie, A. C., ed. 1963. Scientific Change. New York: Basic.

a ay oe oS,

. 1963. “The Evolution of the Physicist’s Conception of Nature.”

Cushing, J. T. 1990. Theory Construction and Selection in Modern
Physics: Scientific American 208 (5): 45—53. :
The S Matrix. Cambridge: Cambridge University Press.
.1977.”Recollections of an Exciting Era.” In Weiner 1977, 109-46.
- 1994a. “A Bohmian Response to Bohr’s Complementarity.” In
Douglas, M. 1975. Implicit Meanings: Essays in Anthropology. London:
Faye and Folse 1994, 57-75.
MeenaOPENae i

Routledge and Kegan Paul.

 aahabiala
338 References
References 339
- 1986. How Institutions Think. Syracuse, N.Y.: Syracuse Univer- Elsasser, W. 1925. “Bemerkungen zur Quantenmechanik freier Electro-
sity Press. nen.” Naturwissenschaften 13:711.
Dresden, M. 1987. H. A. Kramers: Between Tradition and Revolution. Ber-
Erikson, E. 1982. “Psychoanalytic Reflections on Einstein’s Centenary”
lin: Springer. In Holton and Elkana 1982.
Duane, W. 1923. “The Transfer in Quanta of Radiation Momentum to
Everett, H. 1957. “’Relative State’ Formulation of Quantum Mechan.
Matter.” Proceedings of the National Academy of Sciences 9:158—64.
ics.” Reviews of Modern Physics 29:454—62. [Reprinted in Wheeler and

Soe careereeemrianienrsoneiney
Duhem,P. 1991. The Aim and the Structure of Physical Theory. Princeton,
Zurek 1983, 315—23.]
N.J.: Princeton University Press. [First published in 1907.] . Faye, J. 1991. Niels Bohr: His Heritage and Legacy. An Antirealism View of
Durr, O., S. Goldstein, and.N. Zanghi. 1992a. “Quantum Equilibrium
Quantum Mechanics. Dordrecht: Kluwer. .
and the Origin of Absolute Uncertainty.” Journal of Statistical Physics Faye,J., and H. Folse, eds. 1994. Niels Bohr and Contemporary Philosophy.
67 :843-907. a Dordrecht: Kluwer.
- 1992b. “Quantum Mechanics, Randomness, andDeterministic Feuer, L. S. 1963. The Scientific Intellectual: The Psychological and Socio-
Reality.” Physics Letters A 172:6-12: logical Origins of Modern Science. New York: Basic.
. 1996. “Bohmian Mechanics as the Foundation of Quantum Me- Feverabend, P. K. 1960. “Professor Bohm’s Philosophyof Nature.” Brit-
chanics.” In Cushing, Fine, and Goldstein 1996, 21-44.
ish Journal for the Philosophy of Science 10:321—38. [Reprinted in Fey-
Earman,J. 1986. A-Primer on Determinism. Dordrecht: Reidel. erabend 1981b, 219-—45.]
- 1993. “Camap, Kuhn and the Philosophy. of Scientific Method- : . 1970. “Consolation for the Specialist.” In Lakatos and Musgrave
ology.” In Horwich 1993, 9-36. 1970, 197-230.
Easlea, B. 1981. Science and Sexual Oppression. London: Weidenfeld and . 1975. Against Method. London: New Left.
Nicolson. i ’ _ . 1981a. “Niels Bohr’s World View.” In Feyerabend 1981b,247.
Einstein, A. 1905. “Uber die Erzeugung und Verwandlungdes Lichtes 93:
betreffenden heuristischen Gesichtspunkt.” Annalen der Physik 17: . 1981b. Philosophical Papers. Vols. 1 and 2. Cambridge: Cam-
132-48. St bridge University Press.
-—. 1917. “Quantentheorie der Strahlung.” Physicalische Zeitschrift
. 1993. Against Method, rev. ed. London: Verso.
18:121-28. Translated as “On the Quantum Theoryof Radiation,”
. 1995.Killing Time. Chicago: University of Chicago Press.
in van der Waerden 1967, 63-77. Page references are to van der Feynman,R. P. 1976. “Los Alamos from Below: Reminiscences of 1943_
Waerden. 1945.” Engineering and Science 39:11-30.
. 1924. “Quantentheorie des einatomigen idealen Gases.” Sit- . 1980. “Los Alamos from Below.” In Badash, Hirshfelder, ang
zungsberichte der Preuftischen Academie der Wissenschaften zu Berlin, Broida 1980, 105-32.
261-267. . 1985. QED: The Strange Theory of Light and Matter. Princeton,
~. 1949a. “Autobiographical Notes.” Reprinted in Schilpp 1970. N_J.: Princeton University Press.
_ 3-94. Page referencesare to Schilpp. - 1998. The Meaningof It All: Thoughts of a Citizen Scientist. Reaq-
—. 1949b. Reply to Criticisms.” Reprinted in Schilpp 1970, 665-88.
ing, Mass.: Addison Wesley.
Page references are to Schilpp. | -
Feynman,R.P., R. B. Leighton, and M. Sands. 1969. The Feynman Lec-
Einstein, A., and S. Freud. 1933. Why War? Paris: International Institute
tures on Physics. London: Addison Wesley.
of Intellectual Co-operation, League of Nations.
Fine, A. 1986. The Shaky Game: Einstein, Realism and the Quantum Theory,
Einstein, A., B. Podolsky, and N. Rosen: 1935. “Can Quantum-Mechani-
Chicago: University of Chicago Press.
cal Description of Physical Reality Be Considered Complete?” Physi-
. 1989. “Do Correlations Need to Be Explained?” In Cushing and
cal Review 47:777-80.
McMullin 1989, 175-95.
Elkana, Y. 1970. “Helmholtz’s ‘Kraft’: An Illustration of Concepts in
. 1993. “Einstein's Interpretations of the Quantum Theory.” Scj-
Flux.” Historical Studies in the Physical Sciences 2:263-98.
ence i Context 6 (1): 257-74.
. 1977. “The Historical Roots of Modern Physics.” In Weiner
- 1996a. “Science Made Up: Constructivist Sociology of Scientific
1977, 197-265. . Knowledge.” In Galison and Stump 1996.
 Lieie tS SRlcaan ethicsaac Sie
340 References
References 341
. 1996b. The Shaky Game: Einstein, Realism and the Quantu
mTheory, Ghirardi, G. C., A. Rimini, and T. Weber. 1986. “Unified Dynamic
2d ed. Chicago: University of Chicago Press. s
Fine, A., M. Forbes, and L. Wessels, eds. 1991. PSA 1990: for Microscopic and Macroscopic Systems.” Physical Review D 34:
Proceedings of 470-91.
the 1990 Biennial Meeting of the Philosophy of Science Associa
tion. Vol. 2. Ghose, P., and D. Home. 1992. “Wave-Particle Duality of Single-Photon
East Lansing, Mich.: Philosophy of Science Association.
Finkelstein, D. 1965. “The Logic of Quantum Physics.” States.” Foundations of Physics 22 (12): 1435~47.
Transactions of Ghose, P., D. Home, and G.S. Agarwal. 1991. “An Experimentto Throw
the New York Academy of Sciences 25 :621—37.
MoreLight on Light.” Physics Letters A 153:403~6.
- 1987. “All Is Flux.” In Hiley and Peat 1987, 289-94.
Fischhoff, B. 1975. “Hindsight * Foresight: The Effect of Outco - 1992. “An Experiment to Throw More Light on Light: Implica-
me of tions.” Physics Letters A:158:95-99, . ee eR eS
Knowledge on Judgment under Uncertainty.” Journal of
Experimental Gibbins, P. 1987. Particles and Paradoxes. Cambridge: Cambridge Univer-
Psychology: Human Perception and Performance 1:288—99.
Fleck, L. 1979. Genesis and Development of a Scientific sity Press. :
Fact. Chicago: Uni- Gillies, D. 1992a. “Are There Revolutions in Mathematics?” In Gillies
versity of ChicagoPress. [First published in 1935.]
1992b.
Fock, V. 1958. Uspechifziceskich nauk 66:599-602. Translation byFock
, , ed. 1992b. Revolutions in Mathematics. Oxford: Clarendon.
Bohr Manuscript Collection, AHOP. Gombrich, E. H. 1963. Meditations on a Hobby Horse, and Other Essays on
Folse, H. 1985. The Philosophy of Nie’s Bohr. Amsterdam: North Hollan
d. the Theory of Art. London: Phaidon. - td :
. 1989. “Bohr on Bell.” In Cushing and McMullin 1989, 254-71
. Gordon, W.1927. “Der Comptoneffect nach der Schrédingerschen The-
. 1994. “Bohr’s Framework of Complementarity and the Realism
orie.” Zeitschrift fur Physik 40:117.
Debate.” In Faye and Folse 1994, 119-39. i Gross, P. R., N. Levitt, and M. W. Lewis, eds. 1996. The Flight from Science
Forman, P. 1971. “Weimar Culture, Causality, and Quantum Theory,
and Reason. New York: New York Academy of Sciences.
1918-1927: Adaptation by German Physicists and Mathematicia
ns to Guillemin, V. 1968. The Story of Quantum Mechanics. New York:
a Hostile Intellectual Environment.” Historical Studies in the Physica
l Scribner's. _ : :
Sciences 3:1-117. , Gutting, G. 1980: Paradigms and Revolutions: Appraisals and Applications
- 1979. “The Reception of Acausal Quantum Mechanicsin Ger-
of Thomas Kuhn’s Philosophy of Science. Notre Dame, Ind.: University of
many andBritain.” In Mauskopf 1979, 11—50. . Notre DamePress.
Franck, J., and G. Hertz. 1913. “Messung der lonisierungsspannu
ng Habermas, T. 1972. Knowledge and Human Interests. Boston: Beacon.
in verschiedenen Gasen.” Vers. Deutsche Physikalische Gesellschaft 15:
Hacking, I. 1983. Intervening and Representing. Cambridge: Cambridge
34-44.
University Press,
Franklin, A. 1986. The Neglect of Experiment. Cambridge: Cambridge
. 1992a. “The Self-Vindication of the Laboratory Sciences.” In
University Press.
French, A., and P. J. Kennedy, eds. 1985. Niels Bohr: A Centenary Pickering 1992, 29-65.
Volume. . 1992b. “Statistical Language, Statistical Truth and Statistical
Cambridge, Mass.: Harvard University Press.
Friedman, M., and H. Putnam. 1978. “Quantum Logic, Condit Reason:‘the Self-Authentification of a Style of Scientific Reasoning.”
a
ional In McMullin 1992.
Probability, and Interference.” Dialectica 32:305-15.
ialaintaNas a2isaaMAS LaRA aPA

Galison, P. 1987. How Experiments End. Chicago: University of Chicag Hanson,N.R. 1958. Patterns of Discovery. Cambridge: Cambridge Uni-
o versity Press. :
Press.
- 1997. Image and Logic. Chicago: University of Chicago Press. . 1959. “CopenhagenInterpretation of Quantum Theory.” Ameri-
Galison, P., and D. J. Stump, eds. 1996. The Disunity of Science can Journal of Physics 27:1-15. [Revised version of paper delivered at
: Bounda- the meeting of the American Philosophical Association, 1958.]
ries, Contexts, and Power. Stanford, Calif.: Stanford University Press.
Gay, P. 1968. Weimar Culture. New York: Harper and Healey, R. 1989. The Philosophy of Quantum Mechanics: An Interactive In-
Row. terpretation. Cambridge: Cambridge University Press.
Geertz, C. 1976. “From the Native Point of View.” In Basso and
Selby Heilbron, J. L. 1982. “Fin-de-Siécle Physics.” In Bernhard, Crawford,
1976, 221-36.
Gentner, D., and M. Jeziorsky. 1993. “The Shift from Metaph and Sérbom 1982. :
orto Anal- . 1987. “The Earliest Missionaries of the CopenhagenSpirit.” In
ogy in Western Science.” In Ortony 1993, 447-80.
Ullman-Margalit 1987, 201-33.
 342 References
References 345
Heilbron,J. L., and T. S. Kuhn, 1969. “The Genesis of the Bohr Atom.” . 1977. “Remarks about the Uncertainty Principle.” In Price and
Historical Studies in the Physical Sciences 1:211—90. Chissick 1977, 3-6.
Heisenberg, W. 1925. “Quantentheoretische Umdeutung kinematischer
. 1979. Philosophical Problems of Quantum Physics. Woodbridge,
und mechanischer Beziehungen.” Zeitschrift fiir Physik 33:879-93. Conn.: Ox Bow. [Reprint of Philosophical Problems of Nuclear Science,
Translated as “Quantum-theoretical Reinterpretation of Kinematical New York: Pantheon, 1952.] ;
and Mechanical Relations,” in van der Waerden 1967, 261-76. Page
Heisenberg, W., and P. Jordan. 1926. “Anwendung der Quantenme-
references are to van der Waerden. 5 eR 8 chanik auf das Problem der anomalen Zeemaneffecte.” Zeitschriftfiir
- 1926a. “Mehrkérperproblem und Resonanz ‘inder. Quanten- Physik 37 :263-77.
mechanik.” ZeitschriftfiirPhysik 38:411-26. s Heisenberg, W., and W.Pauli. 1929. “Zur Quantendynamik der Wellen-
——. 1926b: “Quantenmechanik.” Naturwissenschaften 14.:989—94. felder.” Zeitschrift flir Physik 56:1-61.
-. 1927a. “Schwankungerscheinungen und. Quantenmechanik.” Hendry, J. 1984. The Creation of Quantum Mechanics and the Bohr-Pauli
Zeitschrift fiir Physik 40:501-6.. is & og <4 Dialogue. Dordrecht: Reidel.
. 1927b. “Uber den anschaulichen Inhalt der quantentheoreti- 4 Hermann, A., K. v. Meyenn, and V. F. Weisskopf, eds. 1979. Wolfgang
schen Kinematik und Mechanik.” Zeitschrift flir Physik 43:172-98. “4
Pauli: Scientific Correspondence with Bohr, Einstein, Heisenberg, A.O.
Translated in Wheeler and Zurek 1983, 62-84. Page referencesare to Vol. 1. Berlin: Springer.
5
Wheeler and Zurek. - : eo: so Hermann, G. 1935. “Die naturphilosophischen Grundlagen der Quan-
ti ——. 1928. “Erkenntnistheoretische Probleme in der Modernen Phy- ' tenmechanik.” Abhandlungen der Friesschen Schule 6:75—152.
#
sik.” In Blum, Diirr, and Rechenberg 1984, 22~28. Hershbach, D. R. 1996. “Imaginary Gardens with Real Toads.” In Gross,
BIST

- 1930. The Physical Principles of Quantum Theory. Chicago: Uni- Levitt, and Lewis 1966. :
versity of Chicago Press.. _ Hessen, B. 1931. “The Social and Economic Roots of Newton's ‘Prin-
. 1931a. “Die Rolle der Unbestimmtheitsrelationen in der mo- cipia.’” In Bukharin 1971. —
SaaS
7

dernen Physik.” Monatshefte fiir Mathematik und Physik 38 :365-72.

Hiley, B. T., and F. D. Peat, eds. 1987. Quantum Implications: Essays in
. 1931b. “Kausalgesetz und Quantenmechanik.” Erkenntnis zu- Honorof David Bohm. New York: Routledge.
*

Sleich Analen der Philosophie 2:172-82. [Reprinted in Blum, Diirr, and Holquist, M. 1990. Dialogism: Bakhtin and His World. London:Routledge.
Rechenberg 1984, 29-39.] Holton, G. 1970. “The Roots of Complementarity.” Daedalus, fall, 1015-
—. 1933. “The Developmentof Quantum Mechanics.” In Nobel Lec- 55. [Reprinted in Holton 1973.]
tures in Physics 1965, 290-301. / . 1973. Thematic Origins of Scientific Thought. Cambridge, Mass.:
. 1934. “Wandlungen in den Grundlagen der exacten Naturwis- Harvard University Press.
senschaft in jiingster Zeit.” Naturwissenschaften 40. Translated as “Re- . 1991. “Quanta, Relativity, and Rhetoric.” In Pera and Shea 1991.
cent Changesin the Foundations of Exact Sciences,” in Heisenberg Holton, G., and Y. Elkana, eds. 1982. Albert-Eiustein: Historical and Cul-
_ 1979, 11-26. Page references are to Heisenberg 1979.
tural Perspectives. Princeton, N.J.: Princeton University Press.
——. 1935. “Questioris of Principle in Modern Physics.” In Heisen-
Home, D., and M. A. B. Whitaker. 1992. “Ensemble Interpretation of
berg 1979, 41-52. . : Quantum Mechanics: A Modern Perspective.” Physics Report 210:
. 1948. “Der Begriff ‘Abgeschlossene Theorie’ in der modernen 224-317.
Naturwissenschaft.” Dialectica 2:331-36.
Honner,J. 1987. The Description ofNature: Niels Bohr and the Philosophy of
- 1952. Philosophical Problems of Nuclear Science. New York: Pan-
Quantum Physics. Oxford: Clarendon.
theon.
Hooker, C. A. 1972. “The Nature of Quantum Mechanical Reality: Ein-
- 1958. Physics and Philosophy. New York: Harper and
Row. stein vs. Bohr.” In Colodny 1972.
- 1967. “Quantum Theory andIts Interpretation.” In Rozental
. 1991. “Physical Intelligibility, Projection and Objectivity: The
1967, 94-108.
Divergent Ideals of Einstein and Bohr.”British Journal for the Philoso-
. 1971. Physics and Beyond. New York: Harper and Row.
phy of Science 42:491—511.
- 1973. “Development of Concepts in the History of Quantum
. 1994. “Bohr andthe Crisis of Empirical Intelligibility: An Essay
Theory.” In Mehra-1973, 264-75. ©
 sideie wens
344 References
References 345
on the Depthof Bohr’s Thought and Our Philosophica
l Ignorance.” Kaiser, D. 1994. “Bringing the Human Actors Back on
In Faye and Folse 1994, 155-74. j Stage: The Per-
sona] Context of the Einstein-Bohr Debate.” British Journal for
Horgan,J. 1991. “Profile: Reluctant Revolutionar the His-
y.” Scientific American tory of Science 27:129-52.
15:14-15.
Kalckar, J. 1967. “Niels Bohr and His Youngest Disciples.” In Rozenta
Horwich, P. 1993. World Changes: Thomas Kuhn and l
the Nature of Science. 1967, 227-40.
Cambridge, Mass.: MIT Press. .° i
Howard, D. 1990. “Einstein and Duhem.”In 4 Klein, M. 1970. “The First Phase of the Bohr-Einstein Dialogue.” His-
Pierre Duhem: Historian and j
torical Studies in the Physical Sciences 2:1—39.
Philosopher of Science, edited by R. Ariev and P. a4
a
Barker. Synthese 83: - 1977. “The Beginnings of Quantum Theory.” In Weiner 1977,
363-84. —_ oe 1+39. *§
- 1993. “Was Einstein Really a Realist?” Perspectives
on Science Klibansky, R., ed. 1958. Philosophy in Mid-Century: A Survey. Florence:
1:204-51.
La NuovaItalia Editrice. ts
- 1994. “What Makes a Classical Concept Classical?
Toward a Knorr-Cetina, K. 1981. The Manufacture of Knowledge: An Essay on the
Reconstruction of Niels Bohr’s Philosophyof Physics.”
In Faye and Constructivist and Contextual Nature of Science. Oxford: Pergamon.
Folse 1994, 201-30. Kochen, S. 1985. “A New Interpretation of Quantum Mechanics.” In
Howard,D., andJ. Stachel, eds. Forthcoming. The Young Einstein:
Ein- Lahti and Mittelstaedt 1985, 151-169: -
stein Studies. ; Koestler, A. 1960. The Watershed. NewYork: Anchor.
Hughes, R.I. Y. 1989. “Bell's Theorem, Ideology and Structural Expla-
Kojevnikov, A. 1987. “Electrodynamics in Matrix Mechanics: Discord
nation.” In Cushing and McMullin 1989, 195—208.
in Interpretation of the Theory”(in Russian). Preprint no. 1. Moscow:
Ising, G. 1926. “A Natural Limit for the Sensibility of Galvanometer
s.” Institute for History of Science and Technology. a
Philosophical Magazine 1:827-34. gS ; -- 1997. “Niels Bohr and the Copenhagen Network in Physics,
Jagger, A. M., and S. R. Bordo, eds. 1989. Gender
/ Body/Knowledge: The ‘Part I.” Manuscript. ,
Effect of Outcome of Knowledge on Judgement under Uncert
ainty. New Kojevnikov, A., and O. Novik. 1989. “Analysis of Information Ties Dy-
Brunswick, N_J.: Rutgers University Press.
namics in -Early Quantum. Mechanics, 1925-1927. Acta Historiae Re-
James, W. 1907. Pragmatism. New York: Longmans,
Green. [Reissued, rum Naturalium necnon Technicarum (Prague) 20:115—59.
Indianapolis: Hackett, 1981.] , Koyré, A. 1943. “Galileo and Plato.” Journalfor the History ofIdeas 4:400-
Jammer, M. 1966. The Conceptual Development of Quantum
Mechanics. 47. [Reprinted in Koyré 1968.]
New York: McGraw-Hill.
. 1957. From the Closed World to the Infinite Universe. Baltimore:
. 1974. The Philosophy ofQuantum Mechanics. New York: Wiley.
Jordan, P. 1927a. “Die Entwicklung der neuen Quant
Johns Hopkins University Press.
enmechanik.” Na- . 1964. Melanges Alexandre Koyré. Vol. 1. Paris: Hermann.
turwissenschaften 15:614—23, 636-49.
. 1968. Metaphysics and Measurement. Cambridge, Mass.: Harvard
. 1927b. “Philosophical Foundations of Quantum Theory.”
Na- University Press. ,
ture 119:566-69.
Kragh, H., 1982. “Erwin Schrédinger.and the Wave Equation: The Cru-
- 1927c. “Reply to Campbell.” Nature 119:779.
cial Phase.” Centaurus 26:154-97.
- 1927d. “Schrodinger, ‘Abhandlungen zur Wellenmechanik’
” - 1990. Dirac: A Scientific Biography. Cambridge: Cambridge Uni-
Die Naturwissenschaften 6.
versity Press.
. 1927e. “Uber quantenmechanische Darstellung
von Quanten- Kramers, H. A. 1924. “The Quantum Theory of Dispersion.” Nature 114:
sprungen.” Zeitschriftfiir Physik 40:661—66.
310-11;
. 1927. “Uber eine neue Begriindung der Quant
enmechanik.” Kramers, H. A., and W. Heisenberg. 1925. “Uber die Streuung von
Zeitschriftfiir Physik 40:809-38.
- 1944. Physics in the 20th Century. New York. Strahlung durch Atome.” Zeitschrift fiir Physik 31:681-707.
Krips, H. 1987. The Metaphysics of Quantum Theory. Oxford: Clarendon.
- 1972. Der Naturwissenschaftler vor der religiésen Frage.
Hamburg: . 1996. “Quantum Mechanics and the Post-Modern in One Coun-
Stalling.
Jordan,P., and E. P. Wigner. 1928. “Uber das Paulische Aquiva try.” Cultural Studies 10 (1): 78-114. .
lentver- Kruger, L., G. Gingerenzer, and M. S. Morgan. 1987. The Probabilistic
bot.” Zeitschriftfiir Physik 47:631-51.
Revolution. Vols. 1 and 2. Cambridge, Mass.: MIT Press.
 iasuioildDalalSeed Dit iiaStsaianoeid ase
346 References
References 347

Kuhn, T. S. 1957. The Copernican Revolution: Planetary Astronomy in the Latour, B. 1987. Science in Action: Howto FollowScientists and Engineers
Development of Western Thought. Cambridge, Mass.: Harvard Univer- through Society. Cambridge, Mass.: Harvard University Press.
sity Press. Laurikainen, K. V. 1988. Beyond the Atom. Berlin: Springer.
. 1959. “The Essential Tension: Tradition and Innovation in Sci- Lazarus, R. S., A. D. Kanner, and S. Folkman. 1980. “Emotions: A
entific Research.” In Taylor 1959. [Reprinted in Kuhn 1977.] Cognitive-Phenomenological Analysis.” In Plutchik and Kellerman
. 1963. “The Function of Dogmain Scientific Research.” In Crom- 1980, 189-243. a
bie 1963, 347-369, 386-395. ea Leary, L. G., ed. 1955. The Unity of Knowledge. New York: Doubleday.
- 1970. The Structure of Scientific Revolutions, 2d ed. Chicago: Uni- London, F. 1926. “Uber die Jacobischen Transformationen der Quanten-
versity of Chicago Press. First edition published in 1962. mechanik.” Zeitschrift fiir Physik 37:915—25.
: - 1971. “The Relations between History and the History of Sci- Ludwig, G. 1968. Wave Mechanics. Oxford: Pergamon.
ence.” Daedalus 100:271-304. Reprinted in Kuhn 1977, 127-61. Page
MacKinnon, E. M. 1977. “Heisenberg, Models and the Rise of the Matrix
references are to Kuhn 1977. — Bee Mechanics.” Historical Studies in the Physical Sciences 8:137—88.
. 1977. The Essential Tension: Selected Studies in Scientific Tradition . 1982. Scientific Explanation and Atomic Physics. Chicago: Univer-
and Change. Chicago: University of Chicago Press. , sity of Chicago Press.
—., 1978. Black-Body Theory and the Quantuii Discontinuity, 1894 — . 1994. ‘Bohr and the Realism Debates.” In Faye and Folse 1994,
1912. New York: Oxford University Press.
4
o® eu : ed
279-302.
. 1987. “What AreScientificRevolutions?” In Kriiger, Gingeren- Maddox,J. 1988. “License to Slang Coperihagen.” Nature 332:581.
zer, and Morgan 1987, 7-22. : Masterman, M. 1970. “The Nature of a Paradigm.” In Lakatos and Mus-
. 1991. “The Roadsince Structure.” In Fine, Forbes, and ‘Wessels grave 1970, 59-90.
1991, 3-13. . Maudlin, T. 1997. “Kuhn Defanged: Incommensurability and Theory
- 1993. “Afterwords.” In Horwich 1993, 311-41. , Choice.” Revue Philosophique de Louvain 94:428—46.
Kuhn, T. S., J. L. Heilbron, P. Forman, and L. Allen, eds. 1967. Sources Mauskopf, H., ed. 1979. The Reception of Unconventional Science. Boulder,
for History ofQuantum Physics: An Inventory and Report. Philadelphia: Colo.: Westview.
American Philosophical Society. Mauskopf, S. In press. “The Chemical Revolution.” In Readers Guide to
Lahti, P., and P. Mittelstaedt, eds. 1985. Symposium on the Foundations of the History of Science, edited by Arne Hessenbruch. London: Fitzroy
Modern Physics. Singapore: World Scientific. Dearborn. Derived from a paper presented at the History of Chem-
, eds. 1987. Symposium on the Foundations of Modern Physics. Sin- istry Workshop, Edelstein Center, Hebrew University of Jerusalem,
gapore: World Scientific. fall 1995.
Lakatos, I. 1970. “Falsification and the Methodologyof Scientific Re- McMullin, E. 1991. “Rhetoric and Theory of Choice in Science.” In Pera
search Programmes.” In Lakatos and Musgrave 1970, 91-196. and Shea 1991, 55-76.
+ Lakatos, I., and A. Musgrave, eds. 1970. Criticism and the Growth of , ed. 1992. The Social Dimension of Science. Notre Dame,Ind.: Uni-
Knowledge. Cambridge: Cambridge UniversityPress. versity of Notre DamePress.
Lakoff, G. 1993. “The Contemporary Theory of Metaphor.” In Ortony Mehra,J., ed. 1973. The Physicists’ Conception ofNature. Chicago: Univer-
Sonataaber Laterstale hd aN N/R Nr FS i alae br backs
1993, 202-52. sity of Chicago Press.
Lanczos, C. 1926. “Ubereine feldmassige Darstellung der neuen Quan- —. 1976. “The Birth of Quantum Mechanics.” Talk given at the
_ tenmechanik.” Zeitschriftfiir Physik 35:812-30. Cern Collogium, Geneva.
Landé, A. 1955. Foundations of Quantum Theory. New Haven, Conn: Yale
Mehra,J., and H. Rechenberg. 1982. The Historical Development of Quan-
University Press. - tum Theory. Vols. 1-5. New York: Springer.
. 1965. “Quantum Fact and Fiction.” American Journal of Physics
Merleau-Ponty, M. 1973. Adventures ofthe Dialectic. Evanston, Il.: North-
33:123-27. western University Press.
. 1967. New Foundations of Quantum Mechanics. Cambridge: Cam-
Merton, R. K. 1957. “Priorities in Scientific Discovery.” American Socio-
bridge UniversityPress. :
logical Review 22:635—-59.
- 1968. “Replies.” Physics Today 21:56.
Merzbacher, E. 1983. “Energetic lon-Atom Collisions: Early Beginnings -
 dai
348 References
References 349
and Recent Advances.” Invited paper at the thirte
enth international - 1945. “Niels Bohr on His 60th Birthday.” Revie
conference onthe Physicsof Electronic and Atomi
c Collisions, Berlin.
ws of Modern
Physics 17:97-101.
Miller, A. I. 1978. “Visualization Lost and Regai
ned: The Genesis of the - 1950. “Die Philosophische Bedeutung der Idee
Quantum Theory in the Period 1913-1927.” In Wechsler. der Komple-
mentaritat.”” In Pauli 1964, vol. 2.
- 1984. Imagery in Scientific Thought. Basel: Birkhauser. - 1954. “Wahrscheinlichkeit und Physik.” Dialectica 8:112—24.
Moll, W. G. M., and H.C. Burger. 1925. “’The Thermo-Rel
ay’ and ‘The . 1955a. “Exclusion Principle, Lorentz Group and Reflection
of
Sensitivity of a Galvanometer andIts Amplificat
ion.’Philosophical Space Time and Charge.”In Pauli 1955b, 30—51.
Magazine 6:624-31. _ , ed. 1955b. Niels Bohr and the Development of Physics. NewYork:
Moore, W. 1989. Schrodinger, Life-and Thought. Cambridge:
Cambridge McGraw-Hill. SPS cee 2
University Press.
. 1964. Wolfgang Pauli: Collected Scientific Papers. Vols. 1 and 2.
Muller, F. M. 1997. “The Equivalence Myth of Quan
tum Mechanics, Edited byR. Kronig and V. F. Weisskopf. NewYork: Wiley..
Part IL.” Studies in the History and Philosophy of Moder
n Physics 28 (2): - 1973. Lectures on Wave Mechanics. Cambridge, Mass.: MIT Press. °
219-47.
Peirce, C. S. 1931-35. Collected Papers. Edited by C. Hartshorne and
Murdoch, D. 1987. Niels Bohirs Philosophy of Physics. Cambr
idge: Cam- P. Weiss. Cambridge, Mass.: Harvard University Press.
bridge University Press. Pera, M., and W. Shea. 1991. Persuading Science: The Art of Scientific
Nersessian, N. T. 1995. “Opening the Black Box: Cognitive Rhetoric. Canton, Maégs.: Science History. _
Science and ; -
History of Science.” Osiris 10:194~-215. Pickering, A. 1982. ‘Interests and Analogies.” In Barnes and Edge 1982.
Newton-Smith, W. H. 1981. The Rationality of Science. Bosto , ed. 1992. Science as Practice and Culture. Chicago: Universityof
n: Routledge.
Nickles, T. 1995. “Philosophy of Science and History of Science.” Chicago Press. .
Osiris :
10:139-64. - 1995. The Mangle of Practice: Time, Agency and Science. Chicago:
Nobel Lectures in Physics, 1922-1941. 1965. Vol. 2. Amst
erdam: Elsevier. . University of Chicago Press.
Ortony, A., ed. 1993. Metaphor and Thought, 2d ed. Cambridge:
Cam- Pitowsky, I. 1994. “George Boole’s Conditions of Possible Experience
bridge University Press. — and the Quantum Puzzle.” British Journal for the Philosophy of Science
Pais, A. 1967. “Reminiscences from the Post-War
Years.” In Rozental 45:95-125. :
1967, 215-27. . Plutchik, R., and H. Kellerman, eds. 1980. Emotion Theory, Research and
- 1982. “Max Born’s Statistical Interpretation of Quantum Me- Experience. New York: Academic.
chanics.” Science 218:1193-98. Popper, K. 1959. The Logicof Scientific Discovery. NewYork: Basic. [First
- 1986. Inward Bound: Of Matter and Forces in the Physical World. published in 1935.]
Oxford: Oxford University Press.
. 1963a. Conjectures and Refutations. London: Routledge.
- 1991. Niels Bohr’s Times, in Physics, Philosophy and Polity. Oxfor . 1963b. “Three Views Concerning Human Knowledge.” In Pop-
d:
Clarendon. per 1963a, 97-119. « * _
Pauli, W. 1926a. “Quantentheorie.” In Handbuch
der Physik, 23:1-278. Price, W. C., and S. Chissick, eds. 1977. The Uncertainty Principle and
Berlin: Springer. [Reprinted in Pauli 1964, 1 :269-—549_]
Foundation of Quantum Mechanics. New York: Wiley. ©
. 1926b. “Uber das Wasserstoffspectrum vom Standpunkt
. der Price, W. C., S. Chissick, and T. Ravensdale, eds. 1973. Wave Mechanics:
neuen Quantenmechanik.” Zeitschrift fiir Physik 36:336-63.
[Trans- The First Fifty Years. New York: Wiley. es
lated as “On the Hydrogen Spectrum from the Stand Point
of the Przibram, K. 1967. Letters on Wave Mechanics. New York: Philosophical
New Quantum Mechanics,”in van der Waerden 1967,
387—415.] Library.
. 1927. “Uber Gasentartung und Paramagnetismus
.” Zeitschrift Putnam, H. 1968. “Is Logic Empirical?” In Cohen and Wartofsky 1968,
fiir Physik 41:81-102. 181-241.
. 1932. ““Review of Born’s and Jordan's Eleme
ntare Quantenme- Rashevsky,N. V. 1926. “Einige Bemerkungen zu Heisenbergschen Quan-
chanik.” Naturwissenschaften 18 :602.
tenmechanik.”Zeitschriftfiir Physik 36:153-58.
- 1933. “Die allgemeinen Principien der Wellenmechanik.
Handbuch der Physik, 24 (1): 83-272. Berlin: Springer.
” In Redhead, M. 1987. Incompleteness, Non-locality and Realism: A Prolegome-
non to the Philosophy of Quantum Mechanics. Oxford: Clarendon.
 350 References
References 351

Reichenbach, H. 1959. Modern Philosophyof Science. London: Routledge Schrédinger, E. 1924. “Bohr’s neue Strahlungshypothese und der Ener-
and KeganPaul. giesatz.”” Naturwissenschaften 12:720-24.
Renneberg, M., and M. Walker, eds. 1993. Science, Technology, and Na- 3 . 1926a. “Quantisierung als Eigenwertproblem 1” Annalen der
tional Socialism. Cambridge: Cambridge University Press. i
4 Physik 79:361—76. Translated in Schrédinger 1927a. Page references
Rorty, R. 1979. Philosophy and the Mirror of Nature. Princeton, NJ.: to Schrédinger 1927a. a
Princeton University Press. . 1926b. “Quantisierung als Eigenwertproblem II.” Annalen der
- 1991. Objectivity, Relativism and Truth. Philosophical Papers. Vol. 1. Physik 79 :489—527. Translated in Schrodinger 1927a. Page references
Cambridge: Cambridge University Press.” to Schrédinger 1927a. -
Rosenfeld, L. 1956. “Reviewof Landé’s Foundations of Quantum Theory.” . 1926c. “Quantisierung als Eigenwertproblem IIL.” Annalen der
' Nuclear Physics 1:133-34.--. an Physik 80:437—90. Translated in Schrédinger 1927a. Page references
. 1961. “Foundations of Quantum Theory and Complementar- to Schrédinger 1927a.
ity.” Nature 190:384—-88. oe a ; . 1926d. “Quantisierung als Eigenwertproblem IV.” Annalen der
. 1967. “Niels Bohiin the Thirties: Consolidation and Extension Physik 81:109—39. Translated in Schrédinger 1927a. Page references
of the Conception of Complementarity.” In Rozental 1967, 114—36. to Schrédinger1927a. -
-1971."Men and Ideas in the History of Atomic Theory.” Archive . 1926e. “Der stetige Ubergang von der Micro- zur Macrome-
for the History of Exact Sciences 7:69—90. at og chanik.” Naturwissenschaften 14:664—66. Translated in Schrédinger
Rozental, §., ed. 1967. Niels Bohr: His Life and Work as Seen by His Friends
——————

1927a. Page references to Schrdédinger 1927a.

and Colleagues. New York: Wiley. oO . 1926f. “Uber das Verhaltnis der Heisenberg-Born-Jordanschen
Ruark, A. E. 1935.“Is the Quantum-Mechanical Description of Physical Quantenmechanik zu der meinen.” Annalen der Physik 79:734—59.
Reality Complete?” Physical Review 48:466-67. - Translated in Schrédinger 1927a. Page references to Schrédinger
Ruger, A. 1987. “Atomism from Cosmology: Erwin Schrédinger’s Work 1927a.
on Wave Mechanics and Space-TimeStructure.” Historical Studies in . 1926g. “Zur Einsteinschen Gastheorie.” Physikalische Zeitschrift
the-Physical Sciences 18:377—401. - 27:95-101. ,
Sanders, J. T. 1987. Niels Bohr: Essays and Papers. Two volumes of Bohr’s . 1927a. Collected Papers. London: Blackie.
writing on epistemological matters, gathered from published papers . 1927b. “’Energieaustausch nach der Wellenmechanik.” Annalen
and the unpublished manuscripts in the Bohr Manuscript Collection, der Physik 83:956—68. Translated in Schrédinger 1927a. Page refer-
AHQP. Translations from Danish by Else Mogensen. Deposited at ences to Schrédinger 1927a.
the Niels Bohr Archive in Copenhagen, the Niels Bohr Library at the . 1935. “Die gegenwartinge Situation in der Quantenmechanik.”
AmericanInstitute of Physics, the Bancroft Library at the University Naturwissenschaften 23:807-12, 812-28, 844-49. Translated in Pro-
of California at Berkeley, the Baillien Library at the University of Mel-
ips

ceedings of the American Philosophical Society 124:323-38; reprinted in

bourne, the Edelstein Center at the Hebrew University of Jerusalem, Wheeler and Zurek: 152-—67. Page references are to Wheeler and
“Si if

the Science Museum Librarv in London,the Institute of Technology Zurek.

Library at the University of Minnesota, the Accademia Nazionale . 1950. “What Is an Elementary Particle?” Endeavour 9 (35):

eh ci casi Sab
della Scienze in Rome, the Widener Library at Harvard University, 109-16.
the Centre de documentationin Paris, the Hillman Library at the Uni- . 1952a. “Are There Quantum Jumps?”British Journalfor the Phi-
versity of Pittsburgh, the Deutsches Museum Library in Munich, the losophy of Science 3:109—23, 233-42.
American Philosophical Society Library in Philadelphia, the Bibli- ——. 1952b. “July Colloquium 1952, AHOP.”In Bitbol 1995, 19-38.
oteca Universitaria at Madrid, and the Mugar Memorial Library at
. 1952c. Science and Humanism. Cambridge: Cambridge Univer-
Boston University.
‘eaeco 8

sity Press.
Schilpp, P. A., ed. 1970. Albert Einstein: Philosopher-Scientist. Library of
. 1952d. “Transformation and Interpretation in Quantum Me-
Living Philosophers, vol. 7. Cambridge: Cambridge University Press. chanics.” In Bitbo] 1995, 39-95.
bisethaivemkiinicinnimesRenan 1 aiEONli

[First published, LaSalle,Ill.: Open Court, 1949.]

. 1954. “Science, Philosophy and the Sensates.” Manuscript draft
Schmitt, C. 1983. Aristotle and the Renaissance. Cambridge, Mass.:
Har- for three William James lectures, AHOP. In Bitbol 1995, 123-49.
vard University Press. . 1955. “The Philosophy of Experiment.” Nuovo Cimento 1:5-15.
 352 References

References 353
- 1958. “Might Perhaps Energy Be a Merely Statistical
Concept?” . 1974. The Compton Effect:
Nuovo Cimento 9:162-70. Turning Point in Physics. Ne
Science History. w York:
Schweber, S. S. 1994. QED and the Men Who MadeIt: Dyson ,
, Feynman, Stump, D. J. 1996. “New Directions in the Philoso
Schwinger and Tomonaga. Princeton, N_J.: Prince
ton University Press. ies.” In Galison and Stump 1996, 443-53. Phy of Science Stud-
Scully, M. O., B. G. Englert, and H. Walther.
1991. “Quantum Optical Suppes, P., ed. 1980. Studies in the Foundati
Tests of Complementarity.” Nature 351:111-16. ons of Quantun; Me ch
Seelig, C. 1954. Albert Einstein: Eine dokumentarische Biogra East Lansing, Mich.: Philosophy of Science Associati anics,
phie. Zurich: on,
Europa. Taylor, C. W., ed. 1959. The Third Univ
ersity of Utah Research Conference
Sentfleben, H. A. 1923. “Zur Grundlagen Quantentheor on the Identification of Scientific Talent. Salt
ie.” Zeitschrift Lake City: University oF
ftir Physik 22:127-56.
Utah Press. i a “
Teller, P1989. “Relativity,
Relational Holism, and the Bell] Inequalities/”
Serwer, D. 1977. “Unmechanischer Zwang: Pauli, Heisenberg
, and the In Cushing and McMullin 1989, 208-24.
Rejection of the Mechanical Atom, 1923-1925.” Histor .
ical Studies in Tolansky, S. 1968. Revolutionin Optics. Harmondsworth: Penguin,
the Physical Sciences 8:189—256.
Tomonaga, S. 1962. Quantum
Shapin, S. 1994. A Social History ofTruth. Chicago: University of Chica Mechanics. Translated by Koshiba.
go sterdam: North Holland. Am-
Press.
Turner, M. 1987. Death Is the Mother of Beauty: Mind, Metaphor, Criticism
Shapin, S., and S. Schaffer. 1985. Leviathan and the Air-P
ump. Princeton, Chicago: University of ChicagoPress..
N.J.: Princeton University Press. “
. , Ullman-Margalit, E., ed. 1987. Science in Reflection. Boston Studies a
Shapiro, A. 1989. “The Evolving Structure of Newtons Theor
y of the Philosophy of Science, vol. 110. The Israel Colloquium:
Color.” Isis 71:197-210. / Studies
in History, Philosophy and Sociology of Science, vol. 3. Dordrecht:

Siliete sha SeRabe:

Shimony, A. 1985. “Reviewof Folse’s The Philosophy ofNiel
s Bohr.” Phys- ValKluwer.
entini, A. 1996. “Pilot-Wa
ics Today 38:109. ve Theory ofFields, Gravit
~ 1989. “Search for a World View Which Can Accommodat ation anq Cos.
e mology.” In Cushing, Fine, and Goldstein 1996, 45-66.
Our Knowledge of Microphysics.” In Cushing and McMullin 1989,
——. 1998. On the Pilot-Wave Theory of
25-38. Classical, Quantum and Sub-
Skinner, Q. 1969. “Meaning and Understandingin the Histo quantumPhysics. Berlin: Springer. .
ry of Ideas.” van der Waerden,B. L., ed. 1967.
History and Theory 8:3-53. Sources of Quantum Mechanics. New
Sklar, L. 1992. Philosophy of Physics. Oxford: Oxford Univer York: Dover. :
sity Press. van Fraassen, B. C. 1980. The Scientific Image. Oxford: Clarendo
Slater, J. 1975a. Solid State and Molecular Theory: A Scientific Biogra n.
phy. . 1981. “A Modal Interpretation of
New York: Wiley. Quantum Mechanic s.” In Bel-
trametti and van Fraassen 1981, 229-58.
—. 1975b. “Wave Mechanicsin the Classical Decade, 1923-1932.” . 1989. “The Charybdis of Realism:
In Slater 1975a. « é Epistemological Implications
of Bell’s Inequality.” In Cushing and McMullin 1989, 97-114.
vat
Sommerfeld, A. 1927. “Zum gegenwartigen Stande der Atomp
hysik.” . 1991. Quantum Mechanics: An
Physikalische Zeitschrift 28 :231-39. Empiricist View, Oxford: Clar-
Stachel, J. 1986. “Einstein and the Quantum: Fifty Years of Strugg endon. .
le.” In van Vleck, J. 1973. “Central.Fields in Two vis-a-vis Three
Colodny 1986, 349-87.
An Historical Divertissement.” In Price, Chissick, and Dimensions:
. 1993. “The OtherEinstein.” In Beller, Cohen, and Renn Ravensdale
1993, 1973, 26-37.
275-91.
Storr, A. 1988. Solitude: A Returnto the Self. New York: Vernon, P. 1970. Creativity. Harmondsworth: Penguin.
Free Press. Vickers, B., ed. 1984. Occult and Scientific
Stuewer, R. 1970. “Non-Einsteinian Interpretations of Mentalities in the Renaissance.
the Photoelectric Cambridge: Cambridge University Press.
Effect.” In Historical and Philosophical Perspectives ofScie
nce. Minnesota Vigier, J. P. 1982. “Non-Locality, Causality, and Aether in Quantum Me-
Studies in Philosophy of Science,vol. 5. Minneapolis: Univer
sity of chanics.” Astronomische Nachrichten 303:55—80.
Minnesota Press.
Vygotsky, L. 1962. Thought and Language.-Cambridg
e, Mass. MITPress,
354 Reterences

Wechsler,J., ed. 1978. On Aesthetics in Science. Boston: Birkhauser.

INDEX
Weinberg,S. 1992. Dreamsof a Final Theory. New York: Pantheon.
Weiner, C. 1977. History of Twentieth Century Physics. New York: Aca-
demic.
Weizsacker, C. F. von. 1985. “Reminiscences from 1932.” In French and
Kennedy 1985, 183-90.
- 1987. “Heisenberg’s Philosophy.” In Lahti and Mittelstaedt
1987, 277-93. . omg
Wentzel, G. 1926a. “Die mehrfach periodischenSysteme in der Quan-
. tenmechanik.”Zeitschriftfiir Physik 37:80—91.
. 1926b. “Eine Verallgemeinerung der QuantenBedingungenfiir
die Zwecke der Wellenmechanik.” Zeitschriftfiir Physik 38:518-29.
Wertheimer, M., ed. 1959. Productive Thinking, New York: Harper.
Wessels, L. 1979. “Schrédinger’s Route to Wave Mechanics.” Studies in
Acausality, 7, 110-16, 281; and Bohm, 199; Bell, J. S., 193, 257, 275; and Bohm, 299,
History and Philosoptiv of Science 10:311-40.
and Bohr, 167, 194~—201; and Born, 59; 306; theorem of, 304
. 1980. “The Intellectual Sources of Schrédinger’s Interpreta- andclassical concepts, 194-99; and con- Black, Max, 248
tions.” In Suppes 1980, 59-76. So textuality, 303; and discontinuity, 195; Bloch,Felix, 31, 71
- 1981. “What Was Born’sStatistical Interpretation?” In Asquith and disturbance, 155-60; and. physical Bloor, D., 309n.3
and Giere 1981, 187—200. * os . reality, 155-60. See also Causality; Inde- Bohm,David, 6, 13-14, 53n.2; and acau-
. 1983..”Erwin Schrédinger and the Descriptive Tradition.” In terminism; Inevitability; Uncertainty sality, 199; and Bell, 299, 306; and Bohr,
principle 256; and classical concepts, 197, 206-10;
Aris, Davis, and Stuewer 1983, 254-78. Addressivity, defined, 308 and complementarity, 204, 206n.8; and
- 1989. “The Way the World Isn’t: Whatthe Bell Theorems Force Adiabatic principle, 45, 47 continuity, 207, and de Broglie, 106n.4,
Us to Give Up.” In Cushing and McMullin 1989, 80-96. AHOP. See Archive for the History of 285; and determinism, 199, 206-10; and
Westfall, R. S. 1985. “Science and Patronage: Galileo.and the Telescope.” Quantum Physics, disturbance, 201; and Feyerabend, 299:
Alchemists, 262
Isis 76:11~30. and hiddenvariables, 209; and indeter-
Analogy: by alchemists, 262; and comple- minism, 199, 206~10; and Kuhn, 300;
Westman,R. S. 1994. “Two Cultures or One? A Second Look at Kuhn’s mentarity, 264; correspondence prin- position and momentum in, 295; and
The Copernican Revolution.”Isis 85:79-115. ciple as, 297 positivism, 193; and quantum jumps,
Wheaton,B. 1983. The Tiger and the Shark: Empirical Roots ofWave-Particle Angular momentum,29 77, and quantum potential, 256; and un-
Duality. Cambridge: Cambridge University Press. Anschaulichkeit, 178; and classical con- certainty, 201, 202; and wholeness, 256
Wheeler, J. A., and W. H. Zurek, eds. 1983. Quantum Theory and Mea- cepts, 113-16, 245; and correspon- Bohmian mechanics, 11
dences, 249-50; and Heisenberg, 67—70, Bohr, Niels: and acausality, 167, 194-201;
surement. Princeton, N.J.: Princeton University Press. ~ 104, 230; and space-time, 115; and and antirealism, 172-73, 184~85; and
Williams, W. C. 1936. Paterson. New York: New Directions. thought experiments, 184. See also Intui- Bohm, 256; and Born, 57, 184, 317: and
Wilshire, D. 1989. “The Uses of Myth, Image, and the Female Body in tion; Visualization Campbell, 135-37; and causality, 166 -—
Re-Visioning Knowledge.” In Jagger and Bordo 1989. Antirealism: appeal of, 176-79, 184-85; 67, 194-201; andclassical concepts,
Wise, N.J. 1987. “How Do Sums Count? Onthe Cultural Origins of Bohr’s version of, 172-73, 184-85; op- 160-67, 179-84, 188, 228, 237, 276; and
position to, 185-87; rhetoric of, 171-90 commonlanguage, 259-62; Comolec-
Statistical Causality.” In Kriiger, Gingerenzer, and Morgan 1987.
Archive for the History of Quantum Phys- ture, 8, 74, 108, 117, 126, 131-32, 139-
. 1993. “Pascual Jordan: Quantum Mechanics, Psychology, Na- ics (AHQP), 2 44, 237, 264; complementarism of, 264:
tional Socialism.” In Renneberg and Walker 1993, 224-54. Aristotelian concepts, 161, 205 and Compton,131-35; and the Comp-
Wittgenstein, L. 1953. Philosophical Investigations. New York: Macmillan. Astronomy, 291-94, 301-2 ton effect, 23; conflicting views of, 172;
Zernike, F. 1926. “Die natiirliche Beobachtungsgrenzeder Stromstarke.” Atomic bomb, 189, 265 and consistency, 273; contradictions in,
WGcasGREeen ee ica

Zeitschrift fiir Physik 40:628-38. ;

Atomic Physics (Born), 49 275; and correspondenceprinciple, 141,
164-65, 248-52, 274; and Darwin,4,
Bakhtin, Mikhail, 323, 324, 325 264; dialogical analysis of, 117-44; and
Balmer formula, 28-29, 31 Dirac, 161; and disturbance, 159, 199,
Belief, and consensus, 311 250, 254; andEinstein, 131-35, 165.27;
 356 Index

Index 357
Bohr, Niels (continued) pers, 40-49, 60n.11; and complemen-
and EPR argument, 145-67; errors of, Churchill, Winston, 267
tarity, 189; and Comptoneffect, 23; on and positivism, 284; of quantum
me-
271; and experiments, 178; asfather fig- continuous processes, 20n.5; difference
Citation, 314-15
chanics, 88
ure, 270; and Feynman, 273; and Fock, Classical idea of causality, 216 SR
quotients, 20; and Géttingen-Copenha- Comptoneffect, 23, 37, 71-73, 89, 118n4,
49; and funding agencies, 270n.2; and
gen physicists, 38; and indeterminism, Classical physics, 297; and acausality, 120, 125, 131-35, 223, 224, 235n18-
Galileo, 191-92; harmonies in thought 194-99; and Anschaulichkeit, 113-16, Conceptin flux, 40-41
59, 61, 220-22; and Jordan, 48n.40; Sea
of, 248-52; and Heisenberg, 7, 8, 71-74,
and matrix theory, 19, 31, 38; on new 245; and Bohm, 197, 206-10; and Bohr, Conjugate variables, 66
76, 120, 131, 138~41, 143, 164, 181, 226, 160~67, 179-84, 188, 228, 237, 276; Consensus, 187-89, 310-11
conceptions, 181; opposition to Schré- ere
249; hero worship, 270-85; Hitchcock
dinger,41, 45, 46; and orbital model, 53; classical mechanics, 80; classical/ quan- Conservationlaws,60, 133,134, 137n.12,
lectures, 158n.17; incompatible voices,
and Pauli, 57n.7, 317n.9; probabilistic tum cut, 196-98; and complementarity, 2159 :
9; inseparability arguments of, 151; on interpretation of, 19, 39~49; revision of
160-67, 237; and correspondenceprin- Consistency:'and Bohr, 273; and comple: ~
introspection, 247; and Kantian ideas, history, 39-40; and Schrédinger, 41, 44- ciple, 160-67; and Heisenberg, 180; 199; mentarity, 204;.andcontradiction, 100; ~
160, 162, 180, 182, 205, 249, 252: and
46, 186, 216-20; and space-time, 60; and
indispensability of, 160-67, 179-84, 172; andDirac, 204; and Heisenberg,
Kramers, 260; language of, 263; and 188, 237, 276; and intuition, 207; limita- 112, 204; and Hilbert, 205; and Weiz-
stationary states, 26, 47— 48;statistical
light quanta, 133; and mathematics, ontology of, 230; transformation theory, tion of, 233; and matrix mechanics, 27; sacker, 204. Pe ah
259-62; and measurement, 162, 236;
45; and wave function, 186: and wave-
of motion, 246; and uncertainty prin- Continuity, 246; and Bohm, 207.
natural interpretations of, 191: and non- ciple, 207; and wave-particle, 237. See Contradiction, 100, 172. See also Comple-
particle duality, 43, 240-41
locality, 152, 275; and objectivity, 159; also Realism; specific concepts mentarity; Consistency.
Born-Pauli prebabilistic theory, 49
and ontology, 152, 166; and operational- Closed theory, and paradigm, 289 Conventionalism, 104
Bose, Satyendra Nath, 230
ism, 147; and ordinary language, Cloud chamber experiments, 93 : Coordinates, problem of, 29
Bose-Einstein statistics, 73, 230
163nn.22, 23; and particle-kinematic in- Cognition, and emotion, 316 Copenhageninterprétation: contradic-
Bothe-Geiger experiments, 55n.5, 60, 63,
terpretation, 139; and Pauli, 141-42, Cold War, 265... tionsin, 2; and Einstein, 319; and fi-
* 112, 120, 125, 131-33, 237
177; personal intuition of, 259, 271 pas Collision phenomena, 40-49, 59 nality, 191-210; and inevitability, 10,
Boyle, Robert, 312
philosopher, 12, 58, 119, 144, 164, 245— ‘Bragg, William Henry, 225
Commonlanguage, 180, 259-62 191-210; mythologyof, 153, 172-73;
47; and physical reality, 120, 153-60; Communication system, 312 and opposition stand, 10-11; philo-
Brahe, Tycho, 293-94, 298, 301-2 ©
polyphonyof, 117~44, 151, 172; and Commutation relation, 25, 66-69, 79-80, sophical sources of, 9; polyphonyof,
Bridgman, P. W., 176, 178-79, 203.6
positivism, 150, 152; and predictability, 87, 104, 113 171-90; and positivism, 205; rhetoric
Brownian fluctuations, 92~94
166; and probability, 216n.7; psychol- Comolecture, 8, 74, 108, 117, 126, 131-32, _ of, 191-210
Bub,J., 106
ogy analogyof, 159-60; and quantum 139-44, 237, 264 The Copernican Revolution (Kuhn), 291-94,
Buchwald,J. Z., 318
of action, 199~200; and relativity, 182~ Complementarity, 6, 8, 11-13, 201; analo- 303
Buddha, 247
83, 260, 261, 272; and Schrédinger,35, gies of, 264; and Bohm,204, 206n.8; and Copernican system, 176, 291—93, 301-2
Burger, Herman, 92-93
37n.29, 120, 122~32, 171, 234-41; and Bohr, 46, 117-44, 160, 163, 189; and Correspondence principle, 13, 53, 55n.5,
space-time, 133; and stationary states, Born, 189; and Campbell, 137; and cau- 128, 144, 310; and Anschaulichkeit, 249—
Campbell, Norman: and Bohr, 135~37;
120, 125, 128, 131; andstatistical inter- sality, 163; and classical concepts, 160— 50; and Bohr, 141, 164-65, 248-52, 274;
and chance, 97-98, 135-37; and com-
pretation, 141; as storyteller, 243-45; 67, 237; and consistency arguments, 204; andclassical concepts, 160-67; classi-
plementarity, 137; and Heisenberg, 7,
and subjectivity, 159; and thought ex- 66, 96-99, 315; and space-time, 96; and
and Dirac, 119; and disturbance, 155- cal/quantum cut, 196-98; and Dirac,
periments, 261; on truth, 272; and un- 60; and Einstein, 135, 319; and EPR ar- 165; as formal analogy, 297; and mathe-
stationary states, 137; and time, 135, 137
certainty, 73, 122, 138, 140, 251; and vi- gument, 155-60; existential messageof, matics, 259; and Pauli, 36, 56n.5; and
Casimir, Hendrik, 273
sualization, 236; and wave theory, 35, 265; and Heisenberg, 74, 187n.11; and quantum electrodynamics, 165; and
Causality: and Bohr, 166-67, 194-201;
122-24, 132, 171, 235-41; and indeterminism, 255; kinematic-dynamic, Sommerteld, 56n.5; and uncertainty
and Born, 59; classical idea of, 216; and
Weizsacker, 274-75; writing difficulty complementarity, 163; and determin-
236; and mathematical formalism, 261— principle, 30, 296-97; and visualization,
of, 263. See also Correspondence prin- ism, 92, 196n.3; and Heisenberg, 66, 99—
62; and measurement, 204, 236; as meta- 249-50
ciple; specific concepts, colleagues i7 phor, 243-67, 320; and mind-body Correspondence theoryof truth, 282
101, 110-16, 195; and hidden variables, 3
Bohr-Kramers-Slater theory, 55n.5, 60 n.11, 197; and Jordan, 92—95; and Kantian i
problem, 264; and noncommutativity, Creativity: dialogical nature of, 2, 106;
83, 93, 125, 131-34; and experimental 86-87; and operationalism, 152; and and Gestalt model, 105; and paradigms,
ideas, 280; and objectivity, 159; and pre-
refutation, 120 Pauli, 142; sex and, 266; and space-time, 304-6
dictability, 166; and probability, 45, 97;
Bohr-Sommerfeld theory, 30 118, 128, 163; andstatistical theory, 189; Cushing,J. T., 49, 106n.4, 193, 196n.2
etdhiaica drome AtEheda redndash

and relativity, 280; and space-time, 118,

Boltzmann, Ludwig, 280n.14, 284 in textbooks, 319; and uncertainty, 245,
163; and visualization, 216, 250. See also
Born, Max, 43, 63, 265; and acausality,
59; Acausality; Determinism; Inevitability
251; and wave-particle duality, 6, 11, Darrigol, O., 18n.2, 52, 86n.17, 88, 90, 101,
adiabatic principle, 45, 47; and Bohr, 57, Chaostheory, 6, 281, 289 12, 120, 189, 223-41; and wholeness, 165, 172n.1, 224, 228n.6, 231, 232, 278,
184, 317; on causality, 59; collision pa- 255-56 297n.19
Chemistry, 300, 302
Completeness, 145; and determinism,94; Darwin, Charles, 4, 4n.1, 13, 34n.20, 264
 358 Index
Index 359

Davisson-Germer experiments, 123-24, Dreimannerarbeit, 18, 29, 31 Bohr-Kramers-Slater theory, 120; cru- G6ttingen-Copenhagenphysicists, 20, 36,
223 Duane, William, 74, 96 cial, 224; emphasis on, 182; and experi- 38, 56, 67, 78, 195, 213, 284; and consis-
DeBroglie, Louis, 18, 43, 106n.4, 285 Duhem,Pierre, 176, 319.12 mental refutation, 120; Hacking on, tency, 111; and indeterminism, 221-22,
De Broglie waves, 39n.30, 41, 123-24, 132 Duhem-Quinethesis, 303 317-18; interpretation of, 9; and in- 281; and inevitability arguments, 111;
Debye, Peter, 31 Du Mond,Jesse, 274 tuitiveness, 109; and quantum theory, and interference of probabilities, 240—
Determinism, 13, 39; and Bohm, 199, 206— Diirr, O., 106n.4 52-53, 114; thought-experiments, 12. 41; and statistics, 35; and wave function,
10; and causality, 92, 196n.3; and com- Dyson, Freeman, 319 184. See also Measurement; Physicai re- 277-78; and wave-particle duality, 239—
pleteness, 94; and Einstein, 165 n.27; and |
ality; Positivism; specific experiments 41. See also specific persons, concepts
fate, 209; and matrix mechanics, 26~28; - ! Earman, John, 302
and present moment, 111; and proba- _Ehrenfest, Paul, 37, 140 Fate, and determinism, 209 Hacking, Ian, 189n.12, 317
bility, 93; and quantum jumips, 92-93; . Eigenfunction /eigenvalue rule, 26, 153 Feminism, 267 Hamiltonian,classical, 130
‘and wave-particle issue, 238. See also Einstein, Albert: and Bohr, 131-35, Fermi, Enrico, 32 — Hanson, Norwood,13, 238; and Heisen-
Causality; Indeterminism; Uncertainty 165n.27, 192; and complementarity, 135, Feyerabend,Paul, 106n.5, 172, 191n.1, 213, berg, 295; and incommensurabilitv,
principle . 319; conflicting voices of, 282; conserva- 221; and anarchism, 304; and Bohm, 293-300; and Kuhn,301; and uncer-
Dialogical approach: and Bohr’s comple- tism of, 8; and Copenhagen philosophy, 299; and Kuhn, 300 tainty relations, 295-96; and
mentarity, 117-44; and creativity, 2, 6, 319; and de Broglie, 32, 39n.30, 43; and Feynman,Richard, 1, 173-74, 204; and visualization, 295
106; emotional intensity in, i44; and his- determinism, 165n.27: on dualism, 100; Bohs, 273 Harmonies, Bohr’s idea of, 248-52
toriography, 3, 307-25; to the historyof andfluctuation phenomena, 76; and Fichte, Johann, 5, 58, 67 Heilbron,J. L., 58, 196, 244, 270
science, 2-3; and lesser scientists, 96— Heisenberg, 278-79; and Kantian ideas, Finality: and discontinuity, 87-90; rhetoric Heisenberg, Werner, 1; agreement with
101;-and paradigmatic holism, 304; and 195, 280; and light quanta, 224; and ma- of, 191-210 Bohr, 143; ambition, 139; and Anschau-
paradigms, 3; and polyphony, 171; and trix mechanical method, 30; mono- Fine, Arthur, 194, 317, 322 . lichkeit, 67-70, 104, 230; and Bohr, 7, 8,
scientific texts, 7 / atomic gas paper, 42-44; and National Finkelstein, David, 14, 198n.4, 306 71-74, 76, 120, 131, 138-41, 143, 164,
Dialogism, 308-25; affect and, 316; and Socialists, 280; and particle kinematic Fluctuation phenomena, 76, 78 181, 226, 249; and Campbell, 7, 66, 96—
yey

lesser scientists, 315; and social con- framework, 143; and physicalreality, Fock, Vladimir, 49, 183 99, 315; and causality, 66, 99-101, 110-—
structivism, 322; as social epistemology, 154-55; and positivism, 52, 155; and Folse, H., 141 : 16, 195; Chicago lectures of, 226-27;
321-22 probability, 61; quantum theoryinter- Formalconsistency, 98 clash with Bohr, 71-74, 139; and classi-
" Dirac, P. A. M., 28, 29, 58, 228; and Bohr, pretation of, 187n.10; and relativity, 105; Formalism,4, 12, 106, 112; and comple- cal physics, 181, 199; on closed theory,
161; and complementarity, 119; and and stationary states, 84; on theory, 66, mentarity, 261-62 288-89; and complementarity, 74,
consistency, 204; and correspondence 97; and uncertaintyprinciple, 143. See Forman,P., 58, 59n.9 187n.11; and Comptonrecoil, 71-73;
principle, 165; generalized theoryof, also Einstein-Podolsky-Rosen (EPR) Fowler, Ralph, 129 and consistency, 112, 204; contradic-
87-88; and Heisenberg, 85-87; and argument Franck, James, 43, 46, 261 tions in, 184; and corpuscular ontology,
probability, 90; and relativity, 90-91; Einstein-Podolsky-Rosen (EPR) argument, Franck-Hertz experiments, 125, 219 120; and correspondence principle, 164,
and Schrédinger, 32-35; and transfor- 6, 9, 285; and Bell, 193, 257, 275, 299, Frank, Philipp, 176 249; dialogicalcreativity, 6; and Dirac,
mation theory, 86, 90 304, 306; and Bohr, 145-67; and com- Franklin, A., 318 85-87; and discontinuity, 73, 89; and
Discontinuity: and discreteness, 214; and plementarity, 155-60; and disturbance, Friedman, M., 106 disturbance, 91, 104, 112, 201; and Ein-
finality, 87-90; and Heisenberg, 73, 89; 146, 200; measurement arrangement for, stein, 278-79; emotional attachmentof,
andintuitiveness, 116; and matrix theo- 146; nonmathematical summaryof, 153; Galileo, Galilei, 191-92, 205, 282, 290n.4 89; and Fichte, 58, 67;fluctuation paper,
rists, 116n.16; and quantum jumps, 74— andoperationalism, 203 Galison, Peter, 172n.1, 322 76, 194; geometric analysis of, 89: and
78, 214-20; and space-time, 88; and un- Electron positions, 20, 55 Geertz, Clifford, 248 G6ttingen-Copenhagenphysicists, 38;
certainty, 89 : Elementary Quantum Mechanics (Born and Geiger, Hans, 132-33. See also Bothe- G6ttingen program,20; and Hanson,
Disturbance, 9; and acausality, 155-60; Jordan), 38 Geiger experiments 295; on hydrogenintensities, 55; and in-
and Bohm,201; and Bohr, 159, 199, 250, Elkana, 40n.31, 314 Gell-Mann, Murray, 107n.6 determinism, 28, 63-64; and inevi-
254; and complementarity, 155—60; and
Elsasser, Walter, 44 Geometry:classical vs. quantum, 20; and tability arguments, 112; and
conservation of momentum, 201—2; and Emotion, and cognition, 316 commutation formula, 87; and Heisen- intuitiveness, 109, 230; and Jordan, 91—
EPR argument, 146, 200; and Heisen- Empiricism, 205 berg, 89; non-Pascalian, 87; within the 96; and Kantian ideas, 195, 199; and
berg, 91, 104, 112, 201; and measure- Entanglement, 13, 152, 255-58 atom, 19. See also Intuition; Space-time; measurementproblem, 67; and National
ment, 156, 254; and physical reality, Entheorizing, 194 Visualization Socialists, 280; and observer, 91, 112,
155-60; and wholeness, 253-54. See also Epistemic virtues, 309n.3 Gestalt model, and creativity, 105 201, 213; on ordinary language, 198; and
Measurement; Uncertainty principle EPR.See Einstein-Podolsky-Rosen (EPR) Ghirardi-Rimini-Webertheory, 6—7, 289 operationalism, 113, 156n.15; and parti-
Donnan,Frederick, 178 argument Goldstein, S., 106n.4, 193 cle-kinematic meaning, 81; and Pauli,
Dopplereffect, 23 Exact science, notion of, 288 Gombrich, Emst, 315 54, 68-70, 79-85; physical interpreta-
Dostoyevsky, Feodor, 323-24 Experiments, 108; and Bohr, 178; and Gordon, William, 37 tion, 89; polyphonyin, 103-16; and
 360 Index
Index 361
Heisenberg, Werner (continued ) indecision on, 61; of measurement,
positivism, 52-53, 113-14, 205; princi- historiography, 304—6; and incommen-
93; probabilistic interpretation, 4; and contextuality of, 303; and disturbance,
pal axis transformation, 31; and proba- surability, 13, 290-94; on normal sci- 156, 254; and Ghirardi-Rimini-Weber
probability, 61, 62; and quantum jumps,
ence, 306; and quantum revolution,
bility, 249; reinterpretation paper, 20, 52; 215; and transformation theory, 63-64; theory, 289; and Heisenberg, 67; inde-
13n.4, 223
and relativity, 115, 197, 199; response to and wholeness, 255. See also Causality; terminacy of, 93; limits of, 95; nonfor~
Jordan, 94—95; Saxonylecture (1932), Determinism; Inevitability; Uncertainty malizability of, 159; and objectivity, 158,
67; and Schrédinger, 32, 34, 37, 67~85, Lakatos, Imre, 314 283; and operationalism, 147, 182; prob-
Indispensability of classical concepts, doc-
161, 317; and Sentfleben, 7, 66, 96, 99— Landé, Alfred, 206, 240-41 lem of, 67; and stationary states, 127;
trine of, 160-67, 179-84, 188, 237, 276
101, 315; and space-time, 66 ~70, 174; Language. See Ordinary language thermalrelay, 92-93; and uricertainty,
Inevitability: and Copenhagen interpreta-
Lao Tse, 247 202. See also Disturbance: Operational-
spectroscopic data, 19; andstatistics, 91, tion, 10, 191-210; and Heisenberg, 112;
Laplacian determinism, 99, 208. ism; Positivism ~ ot
94, 230; on superiority of matrix me- and operationalism, 199-210; rhetoric
chanics, 114; on theory, 52, 57; turning Latour, Bruno, 40, 172n.1, 309, 314 : 2 Merton, Robert; 316
of, 191-210, 213, 237-38; and wave par-
Lavoisier, Antoine Laurent, 300, 302 Merzbacher, Eugen,49 3
point in quantum physics, 19; two audi- ticle duality, 237-38 Leibnitzian harmony, 108
encesof, 184; uncertainty paper,5, 7, 57, Institutional communication, 312
Micro/macrotransition, 30, 70—71,100.
63-64, 65-101, 103-16; andvirtual os-
Light, 133, 224, 233-36. See also Wave Seealso Space-time;specific contexts
Instrumentalism, 203
cillator model, 174n.4; and visualiza-
mechanics Mind-bodyproblem, 264, 283 -
Interference, 63, 229, 230, 240-41. See also
London,Fritz, 29, 32 Moll, Erich, 92-93
tion, 21, 69-70, 112; and wave-particle Disturbance Lorents, Hendrik, 26 Moller’s student, 160
duality, 225~—27. See also Matrix theory; Interpretation, 6, 9, 107, 109-10, 130 Morin, JeanBaptise, 301.22.
specific concepts, colleagues, issues Intuition, 109; Aristotelian, 161; and Bohr, Mach,Ernst, 52; 284 f . Motion, conceptof, 207, 246 _
Helium atom, 53 160; and classical theory, 207; and dis- MacKinnon, E. M., 53, 108, 121, 174n.5 Murdoch,D., 118; 185, 236
Heraclitus, 14 continuity, 116; and Heisenberg, 109, Magneticquantum numbers, 29
Hermann,Grete, 213 230; Kantian, 195; and motion, 207; and Marxist historians, 209 Narrative structure, 11. See also
Hero worship, of Bohr, 270-85 visualization, 180. See also Anschaulich- Mathematical formalism, 4, 259-62 _ Historiography -
Hertz, Gustav, 46 keit; Visualization Matrix theory, 17-49; and Born,19, 31, 38; National Socialists, 176
Hessen, Boris, 209 Ising, Gustaf, 92~93, 96
cautious tone, 31; and classical theory, NewAge, 267
Hidden variables, 59, 261, 279, 289;
27; and commutation relations, 25, 66— Newton,Isaac, 188 ;
Bohm’s analysis of, 209; and causal James, William, 255 69, 79--80, 87, 104, 113; and determin- Nietzsche, Friedrich, 266
laws, 197; and impossibility proof, 213; Jammer, Max, 5, 5n.2, 11
ism, 26-28; double affix in, 98; and Ein- Noncommutativity, 25, 66-69, 79-80, 86—
and space-time, 60 Jordan, Pascual, 18; on Born, 48n.40; and
stein, 30; equivalence of wave theory, 87, 104, 113. See also Uncertainty
Hilbert, David, 205 : causality, 92-95; and Géttingen-Copen- 36; and Fermi, 32; mathematical Nonlinear systems, 281
Historiography: dialogical, 3, 307-25; and hagenphysicists, 38; and Heisenberg, difficulty of, 28-30; and Maxwell’s Nonlocality, 13, 53n.2, 152, 220, 258, 275,
indeterminism, 220-22; Kuhnian, 304- 91-96; Nazi days of, 267; on probabil- equations, 21; nonintuitiveness of, 178; 281. See also Einstein-Podolsky-Rosen
6; and relativism, 290; of science, 172, ity, 18, 62-64; and Schrédinger, 18, 33, not immediately recognized, 19; and (EPR) argument
209 231-32 particles, 18, 22-25; and Pauli, 54, 79; Normalscience, Kuhnian notion of, 306
Hoffding, Harald, 245, 252 Jung, C.J, 257 and physical interpretation, 21, 30; Number, and law, 20n.5
Holism, 13; and paradigms, 288, 302-4.
probabilistic, 22; and quantum jumps,
See also Wholeness Kantian ideas, 185, 203; and Bohr, 160,
25-26, 75, 77; and radiation, 27; re- . Objectivity, 158-59
Holquist, M., 324 162, 180, 182, 205, 249, 252: and causal-
Holton, G., 58, 104n.1
ceived storyof, 3; and Schrédinger, 30- Observation. See Measurement
ity, 280; and Einstein, 195, 280; and Hei-
Howard, Don, 119, 275, 282 39, 178; and space-time, 19, 21-22, 30, Oldenburg, Henry, 312
senberg, 195, 199; and intuition, 195;
33n.20, 34; and stationary states, 25-26, Ontology, 49, 64, 152, 166
Hydrogen atom, 28; and Balmerterms, 31; andrelativity, 205, 280; and Schré- 75; and statistical methods, 19, 27, 31; Operationalism, 56n.5, 178; and Bohr, 147,
and Pauli, 18, 31n.16; Schrédinger’s so- dinger, 282; and space-time, 180
success of, 31, 114; and uncertainty rela- 156; and Bridgman, 178-79; and com-
lution of, 29; structure of, 18 Kepler, Johann, 293~94, 298, 301-2, 313
tions, 296n.14; use of, 18, 31; and wave plementarity, 152; and Einstein-Podol-
Kierkegaard, Soren, 252, 265n.25
Incommensurability, 293-300 mechanics, 3, 24—25, 32, 36, 44. See also sky-Rosen (EPR)challenge, 203; and
Klein, Oskar, 73
Incompleteness, 175, 297 Heisenberg, Werner Heisenberg, 113, 156n.15; and inevi-
Kojevnikov, A., 36n.26, 52, 270n.2
Matter, and radiation, 133 tability, 199~210. See also Measurement;
Indeterminism,6, 26, 48, 171; and Bohm, Koyré, Alexandre, 291 Maxwell's equations, 21 Positivism
199, 206-10; and Born, 59, 61, 220-22; Kramers, Hendrik, 23, 24 n.12, 53, 134, 164,
MeMullin, E., 246 Orbital model, 53
and chaotic systems, 281; and com- 260
plementarity, 255; in flux, 59-64; and Measurement, 5, 69, 107, 144, 212, 293; and Ordinary language, 182, 184; and Bohr,
Kuhn, Thomas,3, 106n.5, 288; and Bohm,
Géttingen-Copenhagen physicists, 221- atomic phenomena, 157—58, 163; and 163nn.22, 23; and commonsense, 180
300; and Duhem-Quinethesis, 303; and
22, 281; and historiography, 220-22;
Bohr, 162, 236; classical quantum cut, Otherness, 309, 323-24
Feyerabend, 300; and Hanson, 301 j;and
196-98; and complementarity, 204, 236; Overdetermination, 309
ait rac
362 Index Index 363

Pais, Abraham, 11, 49 uncertainty paper, 103-16; in novels, Radiation, 27, 133 161; and wave function,6, 26, 34, 47, 48,
Paradigms, 270-—85, 288-306; and anoma- 324 Ramsauer effect, 44n.35, 223 127; and waveontology, 277-78; weight
lies, 304; and closed theory, 289; and Popper, Karl, 172, 313-14 Realism, 104, 219; criterion for, 146, 153; function of, 220. See also Wave mechan-
creativity, 304—6; defined, 304; dialogi- Position momentum,and uncertainty, 85 and Einstein, 154-55; andinseparabil- ics; specific concepts, issues
cal analysis, 3; dispensability of, 5; his- Positivism, 4, 7, 52, 104, 106, 172, 176, 321; ity, 151; and probability, 186; and rheto- Schultz, Betty, 140
torical inaccuracy of, 288; and holism, and Bohm, 193; and Bohr, 150, 152; and ric, 171-90; and Schrédinger, 292 Schweber,S. S., 232
288, 302-4; incommensurate, 13, 290—- completeness, 284; and Copenhagenin- Redhead, M., 106 é Scientific communication, 312
300; interpretationof, 6; penetrability terpretation, 205; and Einstein, 52, 155; Relativism, in history, 290 Scientific institutions, 312
of, 303; and revolution, 304-6 - in flux, 52-59; and Heisenberg, 52-53, Relativity, 195; and Bohr, 182-83, 260, 261, Selection rules, 29
Particle-kinematic interpretation, 79-82, 113-14, 205; and instrumentalism, 203; 272; and causality, 280; and Dirac, 90 — Sentfleben, H. A., 7, 66, 96, 99-101, 315
1990 gs ai physicalreality, 120, 153-60; and Pop- 91; and Einstein, 105; and Heisenberg, Shimony, Abner, 275
Particle theory:.and Born, 43; conceptof, per, 172, 313-14; and Schrédinger, 283- 115, 197, 199; and Kantian ideas, 205, Skinner, B. F., 172
233; deconstruction of, 284; definition 84; and wavetheory, 75. See also Mea- 280; and Schrédinger,278; and subjec- Slater, John, 23, 83, 259
of, 234; and matrix theory, 18, 22~25; surement; Operationalism tivity, 159 Social constructivism, 322
myth of equivalence. 18: ontology, 49, Postmodernists, 12n.3 Resonance phenomena,216 Sociology of knowledge, 310-16
64; and Pauli, 49, 80. See also Wave Practice, theory as, 318 Revolution, and paradigm, 304-6 Sommerfeld, Arnold, 30, 32, 56n.5, 195
mechanics Predictability. See Causality; Determinism Rosenfeld, Leon, 119, 212 Space, and matrix theory, 21
Patterns of Discovery (Hanson), 293-300 Probability. See Statistical theory Royal Society, 312 Space-time, 66; and Anschaulichkeit, 115; in
Pauli, Wolfgang; and Balmer formula, 28; Psychology: of knowledge, 319-17; and Ruark, Arthur E., 149 atom, 45, 52-55, 60, 95, 118n.1, 130; and
and Bohr, 141-42, 177; and Born, 57n.7, . physics analogy, 159-60, 245-47, Rutherford, Ernest, 215 Bohr, 133; and Born, 60; and Campbell,
317n.9; and complementarity, 142; and 264-65 96; and causality, 118, 163; and comple-
correspondence principle, 36, 56n.5; en- Ptolemaic system, 291-92, 301 Schlick, Moritz, 99 mentarity, 118, 128, 163; and disconti-
cyclopediaarticle of, 84, 142; and Gét- Scholasticism, 205 nuity, 88; and electron position, 55, 95;
tingen-Copenhagen physicists, 38; and QED. See Quantum electrodynamics Schopenhauer, Arthur, 257 elimination of classical, 20-21, 56; elimi-
Heisenberg, 54, 79-85; and: hydrogen Quantization:interpretation of, 130; pos- Schrédinger, Erwin, 43; “beat” model, nation ofintuitive, 33n.20; geometric in-
atom, 18, 29; 31n.16; and matrix me- tulate of, 129; rules, 20; and Schré- 32n.19; and Bohr, 35, 37n.29, 120, 122— terpretation of, 87; and Heisenberg, 66—
chanics, 54, 79; and mystical ideas, 257; dinger, 34, 278; and wave-particle issue, 32, 171, 234-41; and Born,41, 44—46, 70, 174; and hidden variables, 60; intra-
operational attitude of, 54; and particle- 232 186, 216-20; bound system in, 221; and atomic, 45, 60, 130; and Kantian ideas,
kinematics, 80; and particle ontology, Quantum electrodynamics (QED): and conservatism, 8, 18; in Copenhagen,37; 180; limits of precision in, 100; and ma-
49; and probabilistic interpretation, 48, correspondenceprinciple, 165; and and Dirac, 32-35; and Einstein—de Bro- trix theory, 19, 21-22, 30, 33n.20, 34;

‘
57, 81, 84, and Schrédinger, 33, 81, 220; wave-particle duality, 232 glie waves, 32; and fluctuation phe- and Planck's constant, 100; and proba-
andstationary states, 84; and transition | Quantum jumps, 40; and Bohm, 77; as nomena, 76; and Géttingen-Copenha- bility, 97; and Schrédinger,25; and sta-
processes, 84; and uncertainty, 79, 84, central issue, 56; and determinism, gen program, 195; and Heisenberg,32, tionary states, 125; statistical meaning
142; and wave function, 35; and whole- 92-93; and discontinuity, 74-78, 214- 34, 37, 67-85, 161, 317; and his oppo- for, 82; and uncertainty, 100; and virtual
ness, 256-57 20; and Franck-Hertz experiments, 219; nents, 74-75; and intuition, 68; James oscillators, 19n.4; and visualization, 124,
Peirce, C. S., 323 and indeterminism, 215; and matrix lectures of, 283; and Jordan, 18, 33, 231- 215; and waves, 127. See also Time
Photoelectric effect, 223 theory, 25-26, 75, 77; objections to, 215; 32; and Kantian ideas, 282; and Mach, Spectral analysis, 215
Photon, 133, 224, 233-36. See also Wave and Planck, 75, 100, 212-14, 216, 279; 284; and matrix theory, 30-39, 178; ori- Spencer, Herbert, 266
mechanics and probabilistic interpretation, 46-49; gin of wave ontology, 277; and particle Spin, 29n.15
Physicalreality, 120, 153-60. See also Mea- and Schrédinger, 41, 126, 217-18; and language, 35; and Pauli, 33, 81, 220; and Stark effect, 29
surement; Positivism visualization, 215; and wave mechanics, philosophical questions, 219, 283; physi- Stationary states: and Bohr, 120, 125, 128,
Planck, Max, 212, 216, 279; Planck’s con- 47 cal interpretation, 33; and Planck’s law, 131; and Born, 26, 47—48; and Campbell,
stant, 100; and Schrédinger, 213; and Quantum numbers, 129-30 213; and positivism, 283-84; and proba- 137; in Comolecture, 131; conceptof,
space-time, 100; and stationary states, Quantum ontology, 108, 193 bilistic ideas, 35, 40; and quantization, 128; and conservation, 137n.12; and
214 Quantum postulate, 129, 215 34, 278; and quantum jumps,41, 126, Einstein, 84; experimental evidence for,
Planck's law, 75, 213 Quantum potential, 256 217-18; and realism, 282; rejection of in- 46; and matrix theory, 25-26, 75; and
Poincaré, Henri, 281, 316 Quantum Theory (Bohm), 206 terpretation, 38; and relativity, 278; and measurement, 127; and Pauli, 84; physi-
Poisson brackets, 86 Quantum theory,old: contradictionsin, resonance phenomena, 216; and Som- cal magnitude, 25; and Planck’s hy-
Political factors, 176, 189-90, 265-67, 280 100; vs. field theoretical macroscopic merfeld, 32; and space-time, 25; and sta- pothesis, 214; and probability, 47-49;
Polyphony: of Bohr, 117-44, 151, 172; of theories, 100 tionarystates, 34n.21, 47, 48, 126-27; and quantum postulate, 215; radiation
Copenhageninterpretation, 171-90; and Quantum Theory and Measurement (Whee- andstatistical ontology, 35, 40, 230; and in, 34n.21; and Schrédinger, 34n.21, 47,
dialogical analysis, 171; in Heisenberg’s ler and Zurek), 212 uncertainty, 80, 238; and visualization, 48, 126-27; and space-time, 125; and
 364 Index
Index 365
Stationary states (continued) Tunneling, 234
88; and Fichte, 58; kinematics, 23: and
transition probabilities, 62; and uncer- Two-slit experiments, 229, 234, 235 n.18, Weinberg, Steven, 5
matrix mechanics, 32, 44; and matrix Wentzel, Gregor, 28, 29, 49
tainty, 131; and virtual fields, 125; and 256
theory, 3; and Pauli, 35; positivist, 75; Wertheimer, M., 105
wave function, 26, 34n.21, 47, 48, 127
Statistical theory, 5, 6, 278; in atom,19;
probabilistic interpretation of, 39~49; Wessels, Linda, 45, 304
Uncertainty ofscience, 1, 201 probability, 35; and quantum jumps,
and Bohr, 141°216n.7; and Born, 19, 39— Uncertainty principle, 1, 201; and Bohm, j 47; Westman, Robert, 291, 301, 303
received storyof, 3; and Schrédinger,6,
49, 221; and bound systems, 221; and 201, 202; and Bohr,73, 122, 138, 140, Weyl, Hermann, 30, 296
26, 34, 47, 48, 127; stationarystate, 26;
causality, 45, 97n.2; and collision phe- 251; and classical theory, 207; and com- Whiggish history, 214,221, =
andstationary states, 26, 34, 47, 48, 127; Wholeness, 13, 252~59; and on 256;
nomena, 59; and complementarity, 189; mutation relation, 79-80, 86-87; and
victory of, 36; and visualization, 124; and complementarity, 255-56; anddis-
and determinism, 93; and Dirac, 90; and complementarity, 245, 251; and corre-
wave ontology, 277-78; wave. packets,
fluctuations, 194; and Heisenberg, 91, spondence principle, 30, 296-97; and turbance, 253-54; and entanglement,
70, 123-24, 277-78. See also Wave-parti- 257~58; holism, 13; and indeterminism,
94, 194, 230, 249; and indeterminism,4, discontinuity, 89; and disturbance ex-
cle duality 255; and Pauli, 256~57; of quantum
61, 62; and interference, 63, 229, 230, planation, 91; and Einstein, 143; and Wave-particle duality, 120, 189; and Bohr,
240~—41; and matrix theory, 19, 22, 27, Hanson, 295-96; and Heisenberg, 103— phenomena, 163--64; whole numbers,
235-41; and classical concepts, 237; and 20n.5 :
31; and Pauli, 48, 57, 81, 84; preestab- 16; and matrix theory, 296n.14; and complementarity, 6, 11, 12, 120, 189,
lished harmony, 61; and quantum measurement, 202; and Pauli, 79, 84; Wien, Wilhelm, a7 186
223-41; and determinism, 238; dilemma Wilson experiments, 20. 93
jumps, 46-49; and realism, 186; and position/momentum,85; probabilistic
for experiments, 228; as foundational Winners, narrative of, 13
Schrédinger, 35, 40, 230; and space- basis of, 78; reappraisal of, 6; and Schré- myth, 227-32; and. Gottingen-Copenha- Wise;N.J., 58
time, 97; and stationary states, 47~49; dinger, 80, 238; and space-time, 100; and gen physicists, 239-41; and Heisenberg, Wittgenstein, Ludwig,293
statistical kinematic interpretation, 141; stationary states, 131; statistical, 78, 113;
225-27; myth of equivalence, 18.3;
terms, 100; and time, 97~98; and trans- and time-energy, 85; and transformation . paradoxof, 225-27; and quantization, Young, Thomas, 229
formation theory, 61; and uncertainty, theory, 86; and uncontrollability, 253; 232; and quantum electrodynamics, 232;
78, 113; wave function, 35 andvisualization, 109; and wave-parti-
. Stern-Gerlach experiment, 8485 space-time description, 127; and uncer- Zeeman effect, 29, 53
cle duality, 238. See also Disturbance, tainty, 238; visualization of, 130. See also
Structure of Scientific Revolution (Kuhn), Heisenberg, Werner Zeitgeist, 313
288 Schrodinger, Erwin; specific concepts, Zeno’s paradoxes, 207
Underdetermination, 309 issues
Stuewer, R., 131 Zernike, Frits, 92, 95-96
Unobservable quantities, 20
Subjectivity, 159 Unobservables. See Positivism
Superposition, 123~24, 126
Synchronicity, 257 Van Fraassen, B. C., 107, 176n.7, 319n.12
Van Vleck, John, 29
Text, openness of, 105-6 Virtualfields, 125
Theory: and positivism, 106; as practice, Virtual interlocutors, 7
318-20. See also specific theories, issues Virtualoscillator model, 19n.4, 23, 24, 53,
Thermal measurement, 92—93 132, 133, 174n.4
Thermodynamics, classical limits in, 95 Visualization, 30, 177; and Bohr, 236; and
Thought experiments, 12, 184, 261 causality, 216, 250; and correspon-
Time: and Campbell, 135, 137; and impre- dences, 249-50; and Hanson, 295;and_-
cision, 85; inside the atom, 136; and ma- Heisenberg, 21, 69-70, 112; and intui-
trix theory, 21n.8, 96-97; as statistical tion, 180; and quantum jumps, 215; and
concept, 97-98; time-energyrelation, Schrédinger, 161; and space-time,124,
82-85; and uncertainty, 85. See also 215; and uncertainty formula, 109; and
Space-time wave-packet, 124; of waves, 130. See also
Totalitarian society, 310 Classical physics; Intuition
Transformation theory, 19; and Born, 45; Von Neumann, John, 18n.3, 213, 261,
and Dirac, 63-64, 86, 90; and indeter- 275n.10. See also Hidden variables
minism, 63-64; and Jordan, 19, 63-64; Von Weizsacker, Carl Friedrich, 119, 204,
and probability, 61; and uncertainty 274-75
principle, 86 Vygotsky, L., 255
Transition probabilities: and Born, 27; and
Pauli, 84; a priori, 62; and stationary Wave mechanics: and Bohr, 35, 122-24, i
states, 62; time of, 84. See also Quantum 132, 171, 235-41; and Born, 186; and :
jumps _ conservation, 133, 134; diagonality of,
baiseshassent