ARTICLE IN PRESS

Studies in History and Philosophy of

Modern Physics 37 (2006) 298–315
www.elsevier.com/locate/shpsb

Heisenberg and the wave–particle duality

Kristian Camilleri
History and Philosophy of Science Department, University of Melbourne, Parkville, Victoria 3010, Australia

Abstract

This paper examines the development and meaning of Heisenberg’s notion of wave–particle
equivalence and the way in which it differs from Bohr’s more widely known notion of wave–particle
complementarity. According to the statistical interpretation of the wave function, developed by Born
and Pauli in 1926, the electron is treated as a particle, though it cannot be assigned a well-defined
position and momentum at a given time. On the other hand, from the vantage point of quantum
electrodynamics developed by Jordan, Klein and Wigner in 1927–1928, the electron is described as a
quantized matter wave in three-dimensional space. Heisenberg brought these two empirically
equivalent approaches together in his 1929 Chicago lectures. Whereas Bohr argued that it was
necessary to use wave and particle descriptions alternatively in different experimental arrangements,
Heisenberg insisted that one could interpret the quantum-mechanical equation of motion in terms
of either a wave ontology or a particle ontology. Clarifying the differences between Bohr and
Heisenberg provides a deeper insight into the divergent views which formed the so-called
‘Copenhagen interpretation’ of quantum mechanics.
r 2006 Elsevier Ltd. All rights reserved.

Keywords: Heisenberg; Bohr; Wave–particle duality

1. Introduction

Bohr’s view of the wave–particle duality of light and matter, encapsulated in his notion
of complementarity in 1927, is generally acknowledged as one of the central doctrines of
the Copenhagen interpretation of quantum mechanics. As Barbara Stepansky puts it: ‘‘The
wave–particle duality has been accepted as a part of quantum mechanics that is set within
the context of complementarity and referred to as the Copenhagen Interpretation’’
(Stepansky, 1997, p. 385). While Bohr’s view has been the subject of much detailed

E-mail address: [email protected].

1355-2198/$ - see front matter r 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.shpsb.2005.08.002
 ARTICLE IN PRESS
K. Camilleri / Studies in History and Philosophy of Modern Physics 37 (2006) 298–315 299

historical and philosophical scholarship (Beller, 1992; Held, 1994; Murdoch, 1987), little
attention has been devoted to alternative views of the wave–particle duality, particularly
amongst physicists who defended the ’orthodox’ interpretation of quantum mechanics.
This is largely because much of the literature on the history of wave–particle duality
assumes that after 1927, Bohr’s idea of wave–particle complementarity was widely
accepted by his colleagues, and in particular by Heisenberg. The historical studies of Klein
(1964), Mehra and Rechenberg (2001, pp. 271–272), Jammer (1974, p. 69) and Stepansky
(1997, p. 384) all rehearse the standard view that after an initial period of disagreement in
1927, Heisenberg accepted Bohr’s notion of wave–particle complementarity.
However, this view has recently been called into question by Mara Beller, who has
argued that ‘‘a genuine unanimity of opinion’’ between Bohr and Heisenberg ‘‘never
occurred’’ (Beller, 1999, p. 226). Though Beller’s work is important in recognizing that
there are key differences between Heisenberg and Bohr on the wave–particle duality,
precisely what Heisenberg’s views on this question were remains the source of much
confusion and ambiguity. While scholars have devoted considerable attention to Bohr’s
notion of complementarity, much less is known about Heisenberg’s view of the
wave–particle duality in quantum mechanics, for which he coined the term wave–particle
equivalence. This is unfortunate because Heisenberg’s view represents a unique and original
perspective on the wave–particle duality of matter. Indeed, Heisenberg’s philosophy of
physics has not received the attention it deserves, despite the fact that he wrote extensively
on the philosophical significance of quantum mechanics. This paper seeks to partly remedy
this omission in the literature by examining the development and meaning of Heisenberg
notion of wave–particle equivalence between 1926 and 1928. I then analyse the way in
which Heisenberg’s own notion of wave–particle equivalence differed from Bohr’s
understanding of the wave–particle duality of matter.
Heisenberg’s concept of equivalence brings together two distinct approaches. The first of
these is the statistical interpretation of the c-function originating in Born’s papers on
collisions in mid-1926, and subsequently developed by Pauli and Heisenberg. According to
this standard interpretation, the electron is a particle (though without a well-defined
position and momentum). The second, lesser known, interpretation is that of quantized
matter waves offered by Jordan, Wigner and Klein in 1927–1928, whereby the electron is a
quantized wave-field (though without a well-defined wave amplitude). Both approaches
turned out to be mathematically equivalent and led to the same results, convincing
Heisenberg that the wave and particle representations of the electron were simply different
ways of describing the same thing. In bringing a new perspective to Heisenberg’s
interpretation of the wave–particle duality, this paper contributes to a deeper under-
standing of Heisenberg’s philosophy of quantum mechanics.
Though Heisenberg often gave the impression that he was in general agreement with
Bohr, their interpretations actually differed quite markedly. Whereas for Heisenberg, the
wave–particle duality rests on the different, but empirically equivalent, transformations of
the quantum-mechanical equation of motion, for Bohr, wave–particle duality arises from
the different ways in which the electron manifests itself in different experimental
arrangements. The significance of this divergence can be seen in the different ways in
which Bohr and Heisenberg understood the task of finding a coherent interpretation of
quantum mechanics. Heisenberg’s approach was focused on the meaning of the abstract
mathematical formalism. Bohr, on the other hand, was more concerned with how to
understand the seemingly contradictory experimental phenomena, which seemed to require
 ARTICLE IN PRESS
300 K. Camilleri / Studies in History and Philosophy of Modern Physics 37 (2006) 298–315

both wave and particle interpretations. An appreciation of the points of divergence

between Heisenberg and Bohr brings to light the often overlooked fact that the so-called
Copenhagen interpretation of quantum mechanics was actually a mishmash of different
and sometimes conflicting views.

2. Is the electron a wave or particle?

The question of how to conceive of the physical reality of the electron was one of the key
issues animating debates on the interpretation of quantum mechanics in 1926. Writing to
Fowler on October 26, 1926, Bohr remarked that Schrödinger’s visit to Copenhagen had
provided the occasion for stimulating discussions, which had ‘‘gradually centred
themselves on the problems of physical reality’’ posed by quantum mechanics (Bohr,
1985, p. 14). Again in a letter to Kronig written just days later, Bohr highlighted the
importance of Schrödinger’s visit, suggesting that it had given ‘‘rise to much discussion
regarding the physical reality’’ of the electron in quantum theory (AHQP, Bohr to Kronig,
28 October 1926). Shortly after these exchanges Heisenberg presented perhaps the clearest
expression of what he saw as the fundamental task of quantum mechanics. In a lecture
delivered in Düsseldorf on 23 September 1926, he declared:
Electrons and/or atoms do not possess that degree of immediate reality that pertains
to objects of daily experience. It is the very subject matter of atomic physics and, with
it, also of quantum mechanics, to investigate the kind of physical reality applicable to
electrons and atoms . . .
[Although until now, electrons have been conceptualised as particles that move in
well-defined orbits in the atom,] indications are that electrons have a similar degree of
reality as do light quanta. . . Here we merely wanted to point out that the
investigation of this typically discontinuous element and that ‘kind of reality’ is the
real problem of atomic physics and therefore the subject matter of all deliberations
on quantum mechanics (Heisenberg, 1926b, p. 989).
Whereas Schrödinger hoped that the atomic system would be visualized as a matter wave
extended in the region of space surrounding the nucleus, the renunciation of the electron
orbit in quantum mechanics suggested to Heisenberg that there might exist an analogy
between electrons and light quanta. This had been anticipated by Pauli in a letter to
Kramers in July 1925 (Pauli, 1979), where he wrote: ‘‘in cases where interference
phenomena are present, we cannot define definite ‘trajectories’ for the light-quanta’’ but
according to quantum mechanics ‘‘neither can one define any such trajectories for the
electrons in an atom’’. To this extent Pauli emphasized that ‘‘it can now be regarded as
proved that light quanta are just as much (and just as little) physically real as electrons’’
(Pauli to Kramers, 27 July 1925, pp. 232–235 [item 97]). The problem nevertheless
remained: just how was this ‘kind of reality’ to be conceptualized?
Though critical of the attempts to develop a wave theory of matter, Heisenberg was
nonetheless convinced that Schrödinger’s wave equation had brought to light the
wave–particle duality of the electron, just as Einstein’s concept of light-quanta had
prompted physicists to consider the wave–particle duality of radiation. In a letter to
Dirac on 26 May 1926, he stated: ‘‘I see the real progress made by Schrödinger’s theory
in this: that the same mathematical equation can be interpreted as point-mechanics in
 ARTICLE IN PRESS
K. Camilleri / Studies in History and Philosophy of Modern Physics 37 (2006) 298–315 301

a non-classical kinematics and as a wave theory according to Schrödinger’’ (Mehra and

Rechenberg, 2000, p. 202). Heisenberg went on to claim: ‘‘I had always hoped that the
solution to the paradoxes of quantum theory could later on be found this way’’. Again, in
his lecture in Düsseldorf in September 1926, Heisenberg (1926b) explained: ‘‘The
extraordinary physical significance of Schrödinger’s results lies in the fact that a
visualizable interpretation of the quantum-mechanical formulas contains both typical
features of a corpuscular theory, as well as typical features of a wave theory’’ (p. 993).
Here, we find Heisenberg earliest references to the wave–particle duality, which finds its
complete clarification in his notion of wave–particle equivalence in 1928.
It may seem somewhat strange, given Heisenberg’s well-documented critical reaction to
Schrödinger’s attempts to develop a wave theory of matter, that in his letter to Dirac in
May 1926 he should concede that the same equation of motion could be interpreted both as
a particle and as a wave theory. Indeed, Heisenberg had initially been enthusiastic about
Schrödinger’s first paper. Writing to Dirac on 9 April 1926, Heisenberg explained that one
could ‘‘win a great deal for the physical significance of the theory’’ from Schrödinger’s
mathematical treatment of the hydrogen atom (Dirac, 1977, p. 131). But in his letter to
Dirac on 26 May, he expressed grave reservations over Schrödinger’s efforts to develop a
‘‘wave theory of matter’’ (Mehra and Rechenberg, 2000, p. 202). In a paper on the many-
body problem in quantum mechanics, Heisenberg explained his critique of the wave
interpretation of quantum theory:

Even if a consistent wave theory of matter in the usual three-dimensional space could
be developed, corresponding to the programme of Einstein and de Broglie, this
would hardly yield an exhaustive description of atomic processes in terms of our
familiar space-time concepts. Precisely in view of the emerging close analogy between
light and matter one is inclined to believe that such a wave theory of matter is no
more a complete description of our atomic experiences than a wave theory of light
provides a complete interpretation of optical experiences (Heisenberg, 1926a, p. 412).

Here, Heisenberg again drew attention to the analogy between light and matter, but did
not argue, as Schrödinger had, that what is required in quantum mechanics is a wave
theory of the electron. Heisenberg emphasized instead that in quantum mechanics we have
no option but to abandon the classical space–time description. The behaviour of electrons,
whether conceived of as waves or as particles, cannot be described in terms of ordinary
three-dimensional space, though in his paper Heisenberg does canvass the possibility that
we may want to ‘‘consider a space whose structure [MaX bestimmung] differs substantially
from the Euclidean one, as representing ‘ordinary’ space’’ (Heisenberg, 1926a, pp.
412–413). In the absence of any such quantum conception of space and time in the
microscopic realm, Heisenberg wrote to Pauli on 23 November 1926 that in quantum
theory, ‘‘one no longer knows what the words ‘wave’ and ‘particle’ mean’’ (Pauli, 1979,
p. 360 [item 148]).
As I will show, by 1928 Heisenberg realized that the interpretation of quantum
mechanics could be pursued in two separate ways, one taking the ‘particle’ and the other
taking the ‘wave’ as the fundamental concept. The realization that these two approaches
lead to the same experimental results was, for Heisenberg, critical to the final resolution of
the wave–particle dilemma. Our analysis begins with the concept of the particle in
quantum mechanics, and the probabilistic interpretation of the wave function.
 ARTICLE IN PRESS
302 K. Camilleri / Studies in History and Philosophy of Modern Physics 37 (2006) 298–315

3. Probability waves and the quantum mechanics of particles

In his criticism of Schrödinger’s attempt at a wave theory of matter in 1926, Heisenberg

emphasized that the c-function did not, in general, represent a wave field in ordinary three-
dimensional space, but rather a system in n-dimensional abstract configuration space
where n represents the number of degrees of freedom. Only in the simple case of the
hydrogen atom, in which there was only one electron, could one interpret the wave
equation in three-dimensional space, neglecting the electron spin. This, Heisenberg felt,
made a mockery of Schrödinger’s emphasis on ‘‘the visualizability of his theory’’ as against
the abstract matrix theory (Pauli, 1979, Heisenberg to Pauli, 8 June 1926, p. 328 [item
136]). In a paper written in June 1926, Heisenberg (1926a) commented: ‘‘As far as I can see,
Schrödinger’s method also does not represent a consistent wave theory of matter in the
sense of de Broglie’’ (p. 412). He again made this the focus of his critique of wave
mechanics in a lecture on quantum mechanics in September 1926:
[I]t is possible to establish a wave equation in the coordinate space of 3-f-dimensions
for the problem of the motion of f-particles, which then completely replaces the
quantum-mechanical problem mathematically. . . [However, it must be emphasized
that] until now one has not generally succeeded in directly connecting the
Schrödinger wave in phase space with the de Broglie wave in ordinary [three-
dimensional] space, which should then be analogous to the light wave. The wave in
q-space has therefore until now only a formal meaning (Heisenberg, 1926b,
pp. 992–993).
The Schrödinger equation, Heisenberg stressed, must not be interpreted as representing
a ‘real’ matter-wave in three-dimensional space. In his June paper, Heisenberg argued that
in his view, ‘‘one of the most important aspects of quantum mechanics is that it is based on
the corpuscular representation of matter’’, though he was quick to add, ‘‘it does not
concern at the same time a description of the motion of the corpuscles in our familiar
space–time concepts’’ (Heisenberg, 1926a, p. 412). Schrödinger was well aware of the
difficulties of interpreting the c-function as a wave. In his second paper, he acknowledged
that when treating ‘‘the problem of several electrons’’ using the wave equation, we must
interpret the c-function ‘‘in configuration space, and not in real space’’ (Schrödinger, 1928,
p. 60). Here, we are forced to concede that the c-function is ‘‘not a function of ordinary
space and time as in ordinary wave problems’’, which ‘‘raises some difficulty in attaching a
physical meaning to the wave function’’.
By mid-1926 Born published two important papers, in which he gave a theoretical
treatment of the problem of a collision of an electron with an atom by using the
Schrödinger c-function (Born, 1926a,b). Born was able to represent the collision as a
scattering of waves, but he interpreted the square of the wave amplitude of the scattered
wave as giving the probability of finding the electron deflected in a particular direction.
Here, Born introduced for the first time the statistical interpretation of the c function. The
development of the statistical interpretation of quantum mechanics and its implication for
wave–particle duality has been the subject of much historical work though little attention
has been specifically devoted to the interpretation developed by Heisenberg (Beller, 1990;
Konno, 1978; Pais, 1982; Wessels, 1980).
Though Born’s work constituted a definite advance in the treatment of aperiodic
processes through wave mechanics, Linda Wessels has rightly pointed out that it was in
 ARTICLE IN PRESS
K. Camilleri / Studies in History and Philosophy of Modern Physics 37 (2006) 298–315 303

fact Pauli who first saw the possibility of interpreting the square of the wave amplitude as
the probability of finding an electron at a definite position in configuration space in the
stationary state of the atom (Wessels, 1980, p. 187). In a letter to Heisenberg on 19 October
1926 Pauli explained that jcðqÞj2 dq denotes ‘‘the probability that in a definite stationary
state of the system the coordinates q of a particle lie between q and q þ dq’’. (Pauli, 1979,
p. 347 [item 143]). Pauli added: ‘‘From the corpuscular standpoint it thus already makes
sense for it to lie in multi-dimensional configuration space’’ (emphasis in original). This
interpretation was to prove particularly important for the development of transformation
theory by Jordan and Dirac (Jordan, 1927a, p. 811). Extending this interpretation to the
case of a single free particle in three dimensions, Heisenberg was able to calculate the limits
of the accuracy with which we can determine the position and momentum of a particle at
any given time. His work marks the first general probabilistic interpretation of quantum
mechanics.
Heisenberg’s 1927 paper on the interpretation of quantum mechanics takes the concept
of the particle as fundamental, though it was to transform the very concept of a particle.
Heisenberg explained that if ‘‘we determine the position q of the electron as q0 with an
uncertainty Dq, then we can express this fact by a probability amplitude which differs
appreciably from zero only in a region of spread Dq near q0 ’’ (Heisenberg, 1983, p. 69).
Writing to Bohr on 19 March 1927, Heisenberg explained that ‘‘one can see that the
transition from micro- to macro-mechanics is now very easy to understand; classical
mechanics is altogether a part of quantum mechanics’’ (AHQP, emphasis in original). The
electron’s position and momentum can only be determined within certain limits of
precision. In fact, as Heisenberg demonstrated, the more accurately we determine the
position of the electron, the less accurately we know the momentum. He was thus able to
derive the well-known uncertainty relation for the position and momentum of a particle:
Dq Dp4h=4p.
Although Heisenberg derived the uncertainty relations using Dirac’s matrix transforma-
tion theory in his 1927 paper, it was clear that one could now reinterpret Schrödinger’s
concept of wave packets. Following Heisenberg’s conception, in July 1927, Kennard
explained ‘‘in the meaning of the new theory, the ‘wave packet’ of Schrödinger becomes
reinterpreted as the ‘probability packet’’’ (Kennard, 1927, p. 326). In considering an
electron in a highly excited state of the atom in his 1927 paper, Heisenberg argued that the
wave packet representing the electron inevitably spreads out in space in the course of its
periodic motion. He went on to explain, ‘‘all positions [of the particle] count as likely (with
calculable probability) that lie within the bounds of the expanded wave packet’’
(Heisenberg, 1983, p. 74). The wave packet, no longer constituted the particle, as it had
for Schrödinger. It represented instead the limited region of space within which the particle
could be located. In other words, the extension of wave packet in space now represented
the indeterminacy in the electron’s position.
This interpretation of quantum mechanics would have radical consequences. As
Heisenberg now explained, only the ‘‘determination of the position’’ of the electron
through the act of observation ‘‘selects a definite ‘q’ from the totality of possibilities
and limits the options for all subsequent measurements’’. In this way, ‘‘the results of
the later measurements can only be calculated when one again ascribes to the electron
a ‘smaller’ wave packet’’. Heisenberg now drew the conclusion that ‘‘every
position determination reduces the wave packet back to its original extension’’
(Heisenberg, 1983, p. 74). In discussions at the Como conference in September 1927,
 ARTICLE IN PRESS
304 K. Camilleri / Studies in History and Philosophy of Modern Physics 37 (2006) 298–315

Heisenberg again emphasized the ‘reduction of the wave packet’ in describing the motion
of a free electron:
This wave packet moves away not only in a straight line in space, but also it spreads
out in the course of time. For a new observation, the wave packet gives the
probability of finding the electron at a determined position. The new observation
itself however reduces the packet to its original magnitude Dq, which makes a
selection from a totality of possibilities and thereby reduces the possibilities for the
future. This continual change of the wave picture at an observation appears to me a
fundamental feature of quantum mechanics. One must just put into practice the
conception of ‘probability waves’. The waves do not have immediate reality, of the
kind we had earlier ascribed to the waves of Maxwell’s theory. One must interpret
them as probability waves and therefore expect a sudden change at every observation
(Bohr, 1985, p. 140 emphasis added).
The idea of a ‘collapse of the wave packet’ in the act of measurement would remain highly
controversial. However, by the late 1920s this interpretation had become widely accepted,
notwithstanding the protestations of physicists such as Einstein and Schrödinger. In
September 1927, Ehrenfest (1927) showed that when a particle is acted on by a force, the
centre of the wave packet moves according to the classical equation of motion for a point-
particle. Kennard (1928) and Darwin (1927) undertook the study of the motion of particles
using wave mechanics in 1927–1928, thus cementing the new interpretation. In their
exposition of the new quantum mechanics at the Solvay conference in October 1927, Born
and Heisenberg explained how the wave–particle dilemma could be resolved from ‘‘the
point of view of the statistical conception of quantum mechanics’’. As they now explained:
‘‘The waves are waves of probability’’ (Born and Heisenberg, 1928, pp. 164–165).
As Heisenberg later explained, while a single electron can be described by a wave
(probability) packet in three dimensions, in general the wave equation of quantum
mechanics describes a system in abstract multi-dimensional configuration space, not in the
ordinary three-dimensional space (Heisenberg, 1958b, p. 43). Friedrich Hund points out,
by 1927, Heisenberg, Pauli and Born ‘‘saw the Schrödinger equation as a convenient form
of expressing the quantum modification of classical particle mechanics’’ (Hund, 1974, p.
177). Agreement, however, was not universal: a number of physicists were not persuaded
that the electron could be described as a ’particle’ or a ’material point’ since it could not be
assigned a well-defined trajectory. In 1931 Schrödinger raised this problem:
On a little reflection it will be clear that the object referred to quantum mechanics in
this connection is not a material point in the old sense of the word. A material point
in that sense is a thing situated at a given place. . . And if it has a given place at any
given moment then surely it must have a definite trajectory, and also, as might be
assumed a definite velocity. However, this may be, quantum mechanics forbids the
conception of a well-defined trajectory. . . We have ceased to believe in the circular
and elliptical orbits within the atom. To speak of electrons. . . as material points and
yet to deny that they have definite orbits appears both contradictory and absurd
(Schrödinger, 1957, p. 72).
Throughout the 1930s Schrödinger, Planck and von Laue continued to argue that the
concept of particle had no physical significance. Quantum mechanics, they argued, could
not be regarded as a final or complete theory (Laue, 1932; Schrödinger, 1934). As
 ARTICLE IN PRESS
K. Camilleri / Studies in History and Philosophy of Modern Physics 37 (2006) 298–315 305

Schrödinger later explained, ‘‘to me giving up the path seems giving up the particle’’
(Archives for the history of quantum physics, 12 April 1955). Born and Heisenberg were
nevertheless adamant that the statistical interpretation of the c-function had brought the
development of quantum mechanics to a satisfactory conclusion. In their joint paper at the
Solvay conference in 1927 they declared: ‘‘we hold the mechanics of quanta to be a
complete theory, whose fundamental physical and mathematical hypotheses are no longer
susceptible to modification’’ (Born and Heisenberg, 1928, p. 178). Though the electron
could not be ascribed a well-defined trajectory, it was still possible, Heisenberg maintained
to speak of the electron as a ‘particle’, whose motion, insofar as one could still speak of its
motion, was described by the wave equation of quantum mechanics.
In 1926 Heisenberg, it will be remembered, had foreshadowed the possibility of both a
(non-classical) particle mechanics and a (non-classical) wave theory. There is more, then,
to Heisenberg’s conception of the physical reality of the ‘waves’ than the probabilistic
interpretation of the wave function. A deeper understanding of Heisenberg’s view of the
wave–particle dilemma is only possible once we examine the early development of
quantum electrodynamics in 1927–1928. As Heisenberg commented in 1933, whereas
‘‘Schrödinger’s theory is concerned with waves in multidimensional space’’, the papers
written by Jordan, Klein and Wigner on quantum electrodynamics showed that it was
indeed possible to develop a wave theory of matter in three-dimensional space
(Heisenberg, 1985, p. 230). It is worth noting here that, notwithstanding the work of
Hund (1974, pp. 177–181), Bromberg (1977), Darrigol (1986), and more recently Mehra
and Rechenberg (2001, pp. 191–232), this phase of quantum electrodynamics has received
relatively little attention from historians of science. Yet as we shall see, for Heisenberg, this
development was of critical importance in resolving the problem of the wave–particle
duality. It is to this that we now turn our attention.

4. Quantized matter waves and wave–particle equivalence

In February 1927 Bohr delivered a paper in Copenhagen, written by Dirac, entitled

‘‘The Quantum Theory of the Emission and Absorption of Radiation’’. It was to have a
huge impact on the development of quantum electrodynamics. In it Dirac proposed a
method for quantizing the electromagnetic wave, which was later termed ‘second
quantization’. By treating the light wave as a ‘quantized wave’, Dirac was able to show
that one could obtain the same results as could be obtained by using ‘‘the light-quantum
treatment’’ (Dirac, 1927, p. 265). As Oliver Darrigol (1986) explains, in Dirac’s paper on
wave quantization, ‘‘one starts from the classical wave picture of electromagnetic
radiation, applies to it the formal rules of quantization, and gets the discontinuous
structure of radiation energy’’ (p. 228). Dirac (1927) himself pointed out that through wave
quantization it was now possible to understand how ‘‘the wave point of view is consistent
with the light quantum point of view’’ (p. 265). A major breakthrough had thus occurred
in the understanding of the wave–particle duality of light.
Dirac’s method was to provide the inspiration for a new advance in the development of a
wave theory of matter. In the summer of 1927 Jordan appropriated the method of second
quantization and applied it to the wave equation for multiple electrons in three-
dimensional space. In July he wrote to Schrödinger, explaining that it had occurred to
him that by quantizing the electron’s wave field, ‘‘a complete theory of light and matter
could be derived in which, as an essential ingredient, this wave field operates in a quantum
 ARTICLE IN PRESS
306 K. Camilleri / Studies in History and Philosophy of Modern Physics 37 (2006) 298–315

non-classical way’’ (Schweber, 1994, p. 34). Jordan made this idea public in a paper on gas
degeneracy in July 1927, in which he explained that it was now possible to represent
electrons, analogously to light-quanta, ‘‘by quantized waves in conventional three-
dimensional space’’ (Jordan, 1927b, p. 473 emphasis in original). Jordan concluded:
a quantum-mechanical wave theory of matter can be developed that represents
electrons by quantized waves in the usual three-dimensional space. The natural
formulation of the quantum theory of electrons will be attained by conceiving of light
and matter as interacting waves in three-dimensional space. The basic fact of electron
theory, the existence of discrete electric particles, appears in this context as a
characteristic of quantum phenomena, namely to be equivalent to the fact that
matter waves appear only in discrete quantum states (Jordan, 1927b, p. 480).

In October 1927, Jordan and Klein (1927) published an important paper in which they
were able to derive the wave equation for more than one particle by using Dirac’s method
of quantization starting from a classical wave equation for electrons in three-dimensional
space. In January 1928, Jordan and Wigner (1928) developed further the idea of quantized
matter waves. It was now possible to formulate the interaction of the electron wave field
with the electromagnetic field in a way ‘‘which avoids the wave representation in the
abstract [multidimensional] coordinate space in favour of a representation by quantum
waves in the usual three-dimensional space’’. An explanation was now available for ‘‘the
existence of material particles . . . similar to [Dirac’s explanation of] the existence of light
quanta’’, namely through the quantization of waves (p. 631).
This crucial breakthrough in the quantum theory of the electron paved the way for a
translation of Schrödinger’s wave equation in configuration space into a wave theory of
matter in three-dimensional space. However, these ‘waves’ were not to be understood in the
classical sense. Just as Heisenberg had quantized the classical equation of motion for a
particle in 1925 (with the result that the particle no longer possessed a well-defined position or
momentum), so Jordan quantized the classical wave equation (with the result that the wave
amplitude no longer assumed a well-defined value). Heisenberg drew the following conclusion
in 1929: ‘‘it follows that there also exist indeterminacy relations for the wave picture, that can
be derived mathematically from the commutation relations of the wave amplitudes’’
(Heisenberg, 1929, p. 494 emphasis in original). As Born (1956) would put it later, when the
three-dimensional waves are quantized, ‘‘the ‘Anschaulichkeit’ is then lost and the statistical
character of the c-function is introduced at a deeper and more abstract level’’ (p. 143).
Although Heisenberg had throughout 1926 rejected the possibility of a classical wave
interpretation of quantum mechanics along the lines set out by de Broglie and Schrödinger,
he responded to the Klein–Jordan theory of quantized matter waves with enthusiasm. We
have already seen that one of Heisenberg’s key objections to Schrödinger’s original wave
interpretation was that the wave equation only gave a representation in configuration
space of many dimensions. As Heisenberg had repeatedly stressed it was not yet possible to
give a de Broglie wave interpretation in ordinary three-dimensional space. Yet Heisenberg
had foreshadowed such a possibility in his earlier discussion of quantum electrodynamics.
His anticipation of Jordan’s work on the quantized matter wave is perhaps most clearly
expressed in a letter to Pauli on 23 February 1927:
I agree very much with your program concerning electrodynamics, but not quite
concerning the analogy: quantum-wave-mechanics: classical [particle] mechanics ¼
 ARTICLE IN PRESS
K. Camilleri / Studies in History and Philosophy of Modern Physics 37 (2006) 298–315 307

quantum electrodynamics: classical Maxwell [wave] theory. That one must quantize
the Maxwell equations to get light quanta and so on à la Dirac, I believe already; but
perhaps the de Broglie waves will later also be quantized in order to obtain charge
and mass and statistics (!!) of electrons and nuclei (Pauli, 1979, p. 376 [item 154]).
We may reasonably infer from this passage that Heisenberg did not share Pauli’s view
that quantum mechanics is inextricably connected to the concept of a point-mass, while
quantum electrodynamics is based on the concept of the electromagnetic field. Heisenberg
entertained the possibility that quantum mechanics could be derived from the quantization
of matter waves. This possibility emerges again in the report he and Born presented at the
Solvay conference in October 1927 (Born and Heisenberg, 1928, p. 144), but was finally
realized only with the publication of Jordan–Klein and Jordan–Wigner papers later in
1927–1928. As Heisenberg explained to Kuhn in an interview in 1963, ‘‘I was somewhat
excited by the idea of quantizing the Schrödinger waves but that was on account of the
problem of interpretation’’ (Archives for the history of quantum physics, 12 July 1963,
p. 4). For this reason, Heisenberg claimed he had ‘‘found these several papers of
Klein–Jordan–Wigner extremely satisfactory’’ (p. 5). The Jordan–Klein paper, in
particular, was deemed ‘‘an extraordinarily important work’’, not only because it paved
the way for quantum electrodynamics, but also because it shed new light on the
wave–particle duality of matter (Archives for the history of quantum physics, Heisenberg
to Bohr, 5 December 1927).
There was, for Heisenberg, an important difference between the Schrödinger waves and
quantized waves of Jordan, Klein and Wigner. The Schrödinger wave equation could only
be interpreted as a probability wave in abstract configuration space, and therefore
presumed the existence of particles. On the other hand, the wave theory of matter
developed by Jordan, Klein and Wigner was not to be interpreted in this way. Here the
wave amplitudes were not to be interpreted as probability amplitudes, but rather as
representing ‘real’ matter-wave fields in three-dimensional space. Thus, from the
perspective of the quantum electrodynamics of Jordan–Klein and Wigner, the concept
of the wave was, ontologically speaking, more fundamental than the concept of the
particle. In his 1955–1956 Gifford lectures Heisenberg explained:
As early as 1928 it was shown by Jordan, Klein, and Wigner that the mathematical
scheme can be interpreted not only as the quantization of particle motion but also as
a quantization of three-dimensional matter waves; therefore, there is no reason to
consider theses matter waves as less real than the particlesyonly the waves in
configuration space (or the transformation matrices) are probability waves in the
usual interpretation, while the three-dimensional matter waves or radiation waves are
not. The latter have just as much and just as little ‘reality’ as the particles; they have no
direct connection with probability waves but have a continuous density of energy and
momentum, like an electromagnetic field in Maxwell’s theory (Heisenberg, 1958b,
pp. 118, 126 emphasis added).

This passage makes it clear that for Heisenberg, the three-dimensional matter waves,
unlike the probability waves in the usual interpretation, were to be understood as ’real’
waves, in the same way that the ‘particle’ in the probabilistic interpretation was understood
as ‘real’. This marked a shift from the position he had taken at the Como conference,
where he argued that the ‘‘waves do not have immediate reality, of the kind we had earlier
 ARTICLE IN PRESS
308 K. Camilleri / Studies in History and Philosophy of Modern Physics 37 (2006) 298–315

ascribed to the waves of Maxwell’s theory’’. Writing to Bohr on 23 July 1928, Heisenberg
declared, ‘‘I now believe that the fundamental questions are completely solved’’ (Archives
for the history of quantum physics), a conclusion which he reaffirmed in his interview with
Kuhn: ‘‘I made some effort to explain to Bohr that these papers of Klein–Jordan–Wigner
were just a very good illustration of what he wanted with the complementarity because
there was a complete symmetry between waves and particles’’. However, as Heisenberg
recalls Bohr at first ‘‘did not feel exactly that way’’ though precisely what Bohr’s views at
this time were is difficult to judge (Archives for the history of quantum physics, 12 July
1963, p. 6). As we shall see below, Bohr’s understanding of wave–particle duality differed
substantially from Heisenberg’s.
With the emergence of the theory of quantized matter waves, it was now possible to see
that one could arrive at precisely the same equation of motion by a quantum modification
of particle mechanics through the statistical interpretation of the Schrödinger equation or
by a quantum modification of three-dimensional wave theory through quantum
electrodynamics. This is the key to Heisenberg’s claim in 1929 regarding ‘‘the complete
equivalence of wave and particle pictures in quantum mechanics’’ (Heisenberg, 1929,
p. 494). In the appendix of his Chicago lectures in 1929, Heisenberg argued that in
quantum theory, ‘‘the particle picture and the wave picture are merely two different aspects
of one and the same physical reality’’ in the sense that ‘‘one and the same set of
mathematical equations can be interpreted at will in terms of either picture’’ (Heisenberg,
1930, pp. 177–178). This is precisely what Heisenberg had anticipated in his letter to Dirac
in May 1926. In his interview with Kuhn in 1963, he explained how he had presented the
problem:
I wrote these lectures which I gave in 1929. There I think it was always done with the
two possibilities. It was not the dualism in the sense that you needed both, it was
rather a dualism in the sense that you may do it either way. You can either start from
the wave picture [Jordan-Klein-Wigner] or you can start from the particle picture
[Born]. Each of the two pictures can be carried through to the end [or quantized] and
will give you correct answers. At the end you discover that, after all, you have done
the same thing, just in a different language. I think that these lectures at least were
clearly written in the sense that they showed not that we have a dualism, but that you
have the two possibilities (Archives for the history of quantum physics, 23 February
1963, p. 20).
As the passage quoted above suggests, Heisenberg did not think that in quantum
mechanics, the physicist is confronted with a wave–particle duality in the sense that there
exist both matter waves and material particles, as for example de Broglie argued in 1927,
but rather the equivalence of two modes of description. As he later put it, whether we take
the electron to be a particle or a wave, in quantum mechanics, ‘‘we have to quantize both
and then they are the same thing’’ (Archives for the history of quantum physics, 12 July
1963, p. 6). In describing the stationary state, for instance, we can employ either the
language of particles, according to which the c-function describes the probability of
finding an electron at a point in space, or alternatively, the language of wave theory,
according to which we speak of the distribution of charge density surrounding the nucleus.
The ’electron’ can therefore be treated as a particle, whose position and momentum can be
determined only within certain limits of accuracy, or conversely, as a spatially extended,
quantized wave field in three-dimensional space.
 ARTICLE IN PRESS
K. Camilleri / Studies in History and Philosophy of Modern Physics 37 (2006) 298–315 309

5. Heisenberg and Bohr: divergent views of wave–particle duality

Heisenberg’s conception of the wave–particle duality was closely linked with the
mathematical physics of Born, Pauli, Jordan and Klein. Bohr’s views, on the other hand,
were shaped more by his preoccupation with experimental phenomena such as the
interference and diffraction of electrons demonstrated by Davisson and Germer in April
1927, which appeared to reveal the wave-like nature of matter. In other words, whereas
Bohr was concerned primarily with the experimental phenomena in his notion of
wave–particle duality, for Heisenberg, the duality was illustrated through the different
transformations of the fundamental equation of quantum mechanics. As is well known, the
wave–particle duality was the subject of intense discussions between Bohr and Heisenberg
in the spring of 1927, during which time Bohr formulated his notion of ‘complementarity’.
The debate between Bohr and Heisenberg in the spring of 1927 is often thought to have
finally convinced Heisenberg of the importance of Bohr’s notion of wave–particle
complementarity in quantum theory (Jammer, 1974, p. 69). However, this is not so. As we
shall see, the wave–particle duality continued to be the source of much disagreement and
misunderstanding well beyond their discussions in 1927.
As Heisenberg later recalled, it emerged in the course of these discussions that ‘‘Bohr
sought to take the duality between the wave picture and the corpuscular picture as the
starting point of the physical interpretation’’ (Archives for the history of quantum physics,
28 February 1963, p. 10). In his later accounts Heisenberg stressed that by the spring of
1927 ‘‘there remained conceptual differences’’ between he and Bohr, though these ‘‘just
referred to the different starting points or to the different ways of expressing things, but no
longer to a different interpretation of the theory’’ (Heisenberg, 1960, pp. 46–47). In 1928
Oscar Klein, who was present during the discussions in Copenhagen, explained that the
dispute between Bohr and Heisenberg had revolved around whether wave–particle
complementarity or discontinuity was more fundamental in quantum theory (Archives for
the history of quantum physics, Bohr’s Scientific Correspondence, Klein to Birtwhistle, 3
May 1928). Heisenberg’s latter to Pauli on 16 May 1927 confirms this account. There
Heisenberg expressed a somewhat critical attitude towards Bohr’s new viewpoint:
[W]e have here discussed quantum theory at length. Bohr plans to write a general
treatise on the ‘conceptual structure’ of quantum theory, from the viewpoint that
‘there exist waves and particles’—if one begins at once with that one can naturally
make everything contradiction free. In respects to this work, Bohr has drawn my
attention to the fact that. . . certain points could be better expressed and discussed in
every detail, if only one begins with a quantitative discussion directly with the waves.
Nevertheless, I am naturally now as before of the opinion that the discontinuities are
the only interesting things in quantum theory and that one can never stress them
enough (Pauli, 1979, pp. 394–395 [item 163]).
While a compromise of sorts was reached by Pauli in reconciling the views of Bohr and
Heisenberg in May 1927, the ‘note in proof’ added to Heisenberg’s paper on the
uncertainty relations did not, as many scholars have argued, signify that the two physicists
had reached complete agreement (Stepansky, 1997, p. 384). As Heisenberg would later
recall: ‘‘The main point was that Bohr wanted to make this dualism between waves and
corpuscles as the central point of this problem and to say, ‘That is the centre of the whole
story, and we have to start from that side of the story in order to understand it’’’ (Archives
 ARTICLE IN PRESS
310 K. Camilleri / Studies in History and Philosophy of Modern Physics 37 (2006) 298–315

for the history of quantum physics, 25 February 1963, p. 18). Heisenberg had a different
view: ‘‘for me it was clear that ultimately there was no dualism’’ (in the sense that it was
necessary to employ both wave and particle concepts in the description of one and the
same object), and ‘‘therefore I was always a bit upset by this tendency of Bohr of putting it
into a dualistic form’’ (Archives for the history of quantum physics, 5 July 1963, p. 11).
While Heisenberg’s later writings may give the impression he had reconciled his differences
with Bohr, a closer reading of their respective texts tells a different story.
In the paper on complementarity published in 1928, Bohr was adamant that in
confronting the wave–particle duality in quantum theory ‘‘we are not dealing with
contradictory but with complementary pictures of the phenomena, which only together
offer a natural generalization of the classical mode of description’’ (Bohr, 1928, p. 581
emphasis added). To this extent, Bohr argued that it was necessary to use both wave and
particle concepts in quantum theory. Heisenberg, however, seems to have adopted a
different view. In 1929, he pointed out that while it was certainly possible to determine the
limits of the concept of a particle in quantum theory by recourse to the concept of wave
packets, he maintained it was also possible to do so ‘‘without explicit use of the wave
picture’’. The uncertainty relations for the electron’s position and momentum, Heisenberg
argued, ‘‘are readily obtained from the mathematical scheme of quantum theory and its
physical interpretation’’ (Heisenberg, 1930, pp. 15–16). There is no need to invoke the
wave picture of matter.
A closer examination of Heisenberg’s later recollections of his discussions with Bohr on
this point lends further support to the view that Heisenberg never accepted Bohr’s
complementarity argument. In his interview with Kuhn, Heisenberg conceded that in his
conversations with Bohr he had agreed ‘‘it was convenient also to speak about waves’’ in
describing the motion of electrons, though he maintained, ‘‘it was not essential to do it’’
(Archives for the history of quantum physics, 25 February 1963, p. 20 emphasis added).
One could just as easily restrict oneself to speaking about the motion of a particle in
quantum mechanics. Indeed, Heisenberg explained that in actual fact he had never fully
accepted the view that it was absolutely necessary to use both wave and particle concepts,
though he conceded that it was perhaps useful for physicists to think in these terms. He
recalled the critical period in 1927 during which time he came to a kind of compromise
with Bohr:
I could perhaps say that at that time I had understood that it doesn’t do any harm to
my own explanation if I do it that way. I felt, ‘Well, one might also do it that way.
Why not?’ Therefore I didn’t want to protest too strongly against it. I said, ‘All right,
it may be of some help to play always between both pictures’. For me the essential
point was that I had understood that by playing between the two pictures, nothing
could go wrong. So I didn’t object to playing with both pictures. At the same time I
felt that it was not necessary. I would say it was possible but not necessary (Archives
for the history of quantum physics, 25 February 1963, p. 21).
While Heisenberg often seems to have endorsed Bohr’s view of wave–particle duality in
the 1930s, substantial differences remained that were never fully resolved. In Heisenberg’s
hands, the wave–particle complementarity was transformed from a general philosophical
viewpoint into a pedagogical principle. He told Kuhn in 1963: ‘‘For explaining the gamma
ray microscope to physicists, it was useful to play between both pictures. That I could see.
But it was not absolutely essential’’ (Archives for the history of quantum physics,
 ARTICLE IN PRESS
K. Camilleri / Studies in History and Philosophy of Modern Physics 37 (2006) 298–315 311

25 February 1963, p. 21). In Heisenberg’s mind it was possible to speak exclusively in terms
of light quanta or alternatively in terms of electromagnetic waves. In this way, beneath the
veneer of agreement, Bohr and Heisenberg held quite different views of the dual nature of
light and matter.
Aside from the question of whether or not it was necessary to use both wave and particle
concepts in quantum theory, Heisenberg also seems to have originally misinterpreted Bohr
on the manner in which the wave and particle concepts were to be used together in a
complementary description. Describing Bohr’s analysis of the gamma-ray microscope
thought-experiment in his 1929 lectures, Heisenberg again emphasized ‘‘simultaneous use is
made of deductions from the corpuscular and wave theories of light’’ (Heisenberg, 1930,
183, pp. 22–23). Such a view, however, is at odds with the view Bohr expressed at this time.
In his letter to Einstein on 13 April 1927 on the wave–particle duality of light, Bohr argued
that the concept of light quanta could be brought ‘‘into harmony with the consequences of
the wave theory of light’’ only because ‘‘the different aspects of problem never appear at the
same time’’ (Bohr, 1985, p. 22 emphasis added). In later years Heisenberg seems to have
changed his view, arguing instead that for Bohr, wave and particle pictures were
‘complementary’ to the extent that they were to be employed in mutually exclusive
experimental conditions. Yet, even here, Heisenberg remained equivocal, restricting
himself to the more moderate view that for certain experiments ‘‘it might be more
convenient to imagine that the atomic nucleus is surrounded by a system of stationary
waves’’ rather than to visualize the electron as a particle orbiting the nucleus (Heisenberg,
1958a, p. 40 emphasis added).
Though much has been made of Bohr’s notion of wave–particle complementarity, Bohr
actually devoted very little attention to it in his published writings, and what little he did
write on the subject is ambiguous. Bohr’s writings, as many scholars will attest, are
extremely elusive, and there is no agreement on how to interpret him. It is widely accepted
that after 1927, Bohr’s ‘complementarity’ interpretation rested on the view that both
‘‘wave’’ and ‘‘particle’’ pictures were necessary for a full account of experimental
phenomena. However, on some occasions Bohr appears to have argued it was not possible
to directly observe ‘light quanta’ or ‘matter waves’, and to this extent both must be
regarded merely as ‘‘symbols helpful in the formulation of probability laws’’ (Bohr, 1932,
p. 370). In a seldom quoted lecture in 1930, Bohr expressed this view:
In this sense, phrases such as the ‘corpuscular nature of light’ or the ‘wave nature of
electrons’ are ambiguous, since such concepts as corpuscle and wave are only well
defined within the scope of classical physics, where, of course, light and electrons are
electromagnetic waves and material corpuscles respectively (Bohr, 1932, p. 370).
Bohr presented a similar view in his lecture on ‘Maxwell and modern theoretical physics’
delivered in 1931, in which he argued that even in quantum mechanics we must not forget
that ‘‘only the classical ideas of material particles and electromagnetic waves have a field of
unambiguous application, whereas the concepts of photons and electron waves have not’’
(Bohr, 1931, p. 691). Dugald Murdoch, whose study of Bohr’s philosophy represents by far
the most thorough investigation of Bohr’s views on this question, argues that these
passages, while seldom referred to, provide us with a deeper insight into Bohr’s view of
wave–particle duality. Murdoch concludes that the idea that the electron is fundamentally
a ‘particle’ and that light is fundamentally a ‘wave’ represents Bohr’s most carefully
considered position on the wave–particle duality—a position which finds its clearest
 ARTICLE IN PRESS
312 K. Camilleri / Studies in History and Philosophy of Modern Physics 37 (2006) 298–315

expression in Bohr’s writings after 1929 (Murdoch, 1987, pp. 67–79). If we accept
Murdoch’s reading of Bohr, which draws on both his published and unpublished writings,
we can conclude that Bohr and Heisenberg held sharply diverging views on the ‘reality’ of
waves and particles.
Murdoch’s reading of Bohr is supported by the accounts given by two physicists with
whom Bohr was closely acquainted with—Oscar Klein and Léon Rosenfeld. In an
interview with Kuhn, Klein recalled that in discussions on complementarity in 1927, Bohr
emphasized that it was necessary to use the concept of matter waves in describing
electrons, but he always did so ‘‘with the knowledge that these were not waves in the literal
sense, and he pointed out very strongly also that in the literal sense one might say that
electrons are particles and that the electromagnetic waves are waves’’ (Archives for the
history of quantum physics, 16 July 1963, p. 7). Rosenfeld, who worked closely alongside
Bohr during the 1930s, has also argued that in considering the wave–particle duality of
light, Bohr was inclined to see ‘‘the electromagnetic field as being in some sense more
fundamental than the photon concept’’. While there is general agreement that Bohr held
such a view prior to 1927, Rosenfeld contends that: ‘‘This is a point of view that Bohr
never abandoned’’ (Rosenfeld, 1979, p. 690). According to Rosenfeld, Bohr never accepted
the symmetry of wave and particle descriptions of matter either—in the case of the electron
he held the particle concept to be more fundamental, ‘‘whereas the wave aspect is the
symbolic one’’ (p. 700). The accounts provided by Klein and Rosenfeld are significant, not
only because they give us reason to question the standard view of complementarity, but
also because they attribute to Bohr a view which stands in sharp contrast to Heisenberg’s
notion of wave–particle equivalence.
While there remains some conjecture as to precisely what Bohr’s views were concerning
the ‘reality’ of particles and waves, it is clear that he never attributed the same importance
to the Jordan–Klein–Wigner papers on wave quantization as Heisenberg did. In 1955,
Heisenberg credited Jordan, Klein and Wigner with having ‘‘demonstrated for the first
time the complete equivalence of wave and particle pictures in the quantum theory’’
(Heisenberg, 1955, p. 15). Indeed, by the 1950s Heisenberg would reinterpret Bohr’s
original notion of wave–particle complementarity through his own quite distinct notion of
wave–particle equivalence. In his Gifford lectures in 1955–1956, Heisenberg again argued,
‘‘the dualism between two complementary pictures—waves and particles—is also clearly
brought out’’ by the possibility of interpreting the quantum-mechanical formalism in two
different ways. While the Schrödinger wave equation is normally written to express
the probability of finding ‘‘the co-ordinates and the momenta of the particles’’ in
n-dimensional configuration space, ‘‘by a simple transformation it can be rewritten to
resemble a wave equation for an ordinary three-dimensional matter wave’’. The fact that
we can regard the electron as either a wave or particle, Heisenberg insisted, ‘‘does not lead
to any difficulties in the Copenhagen interpretation’’ (Heisenberg, 1958b, pp. 50–51).
In contrast with Bohr’s notion of complementarity, which centred on the ‘experimental
object’, Heisenberg did not take the view that the electron sometimes manifests itself as a
particle, and other times as a wave. Rather, according to Heisenberg, the electron can
always be described formally as either a wave or a particle, though neither is to be
understood literally in a classical sense. Rather than use both classical pictures in a
complementary description, or in mutually exclusive experimental arrangements, Heisen-
berg argued that one could interpret the quantum equation of motion for an electron either
as a quantized wave in three-dimensional space or as describing the quantized motion of a
 ARTICLE IN PRESS
K. Camilleri / Studies in History and Philosophy of Modern Physics 37 (2006) 298–315 313

particle. For Heisenberg, the emergence of the new quantum electrodynamics gave formal
expression to what he termed the ‘‘complete equivalence of wave and particle pictures in
quantum theory’’. The distinction to be drawn between Heisenberg’s equivalence and
Bohr’s complementarity was disguised by the fact that in later years Heisenberg frequently
used the term ‘complementarity’ in ways that brought it closer to his own notion of
equivalence.

6. Conclusion

Although the Copenhagen doctrine of wave–particle duality is virtually synonymous

with Bohr’s notion of complementarity, Heisenberg pursued his own interpretation of the
wave–particle duality between 1926 and 1928, which reflected both the probabilistic
interpretation of the c-wave, to which he himself had contributed decisively through the
notion of the probability packet, and the ‘quantization’ of the three-dimensional wave
carried out by Dirac, Jordan, Klein and Wigner in 1927–1928. According to Heisenberg,
these two interpretive frameworks constituted the two sides of the wave–particle dilemma.
In the first case, the wave-like phenomena such as diffraction and interference could be
explained by treating electrons as particles. In the second case, the corpuscular nature of
electrons is a manifestation of the quantization of matter waves. As a consequence,
quantum mechanics could be interpreted either way, through two quite different
ontologies.
Heisenberg’s view of wave–particle duality is significant for two reasons. First, because it
offers an original and unique insight into the problem of wave–particle duality, which to
this day stands as one of the central paradoxes of quantum mechanics. And secondly, as it
turns out, Heisenberg’s notion of wave–particle equivalence differed quite markedly from
Bohr’s notion of wave–particle complementarity. This divergence between the two men
calls into question the standard historical view that after a brief period of intense
disagreement in spring 1927, Heisenberg accepted Bohr’s complementarity argument.
It is true that Bohr and Heisenberg both accepted some form of wave–particle duality.
However, Bohr’s emphasis on the complementary use of wave and particle descriptions
arose from his preoccupation with the experimental object of quantum mechanics. The
electron exhibits both wave-like phenomena of interference and diffraction, and particle-
like phenomena in different experimental situations. Heisenberg, on the other hand, saw
the possibility of interpreting the object of quantum mechanics as a wave or as a particle,
because of the different transformations of the quantum-mechanical equation of motion.
To this extent, his understanding of the duality arose from the possibility of interpreting
the mathematical formalism in two different, but empirically equivalent, ways. By
highlighting the hidden disagreements between Bohr and Heisenberg, we can see more
clearly the different viewpoints which remained in tension within the Copenhagen
interpretation.

References

Archives for the history of quantum physics. Philadelphia: American Philosophical Society.
Beller, M. (1990). Born’s probabilistic interpretation: A case study of ‘concepts in flux’. Studies in the History and
Philosophy of Science, 21, 563–588.
 ARTICLE IN PRESS
314 K. Camilleri / Studies in History and Philosophy of Modern Physics 37 (2006) 298–315

Beller, M. (1992). The birth of Bohr’s complementarity: The context and the dialogues. Studies in the History and
Philosophy of Science, 23, 147–180.
Beller, M. (1999). Quantum dialogue: The making of a revolution. Chicago: University of Chicago Press.
Bohr, N. (1928). The quantum postulate and the recent development of atomic theory. Nature (Suppl. 5), 580–590.
(Reprinted in Bohr (1985, pp. 147–158)).
Bohr, N. (1931). Maxwell and modern theoretical physics. Nature, 128, 691–692 (Reprinted in Bohr (1985,
pp. 357–360)).
Bohr, N. (1932). Chemistry and the quantum theory of atomic constitution. Journal of Chemical Society, 349–384.
(Reprinted in Bohr (1985, pp. 371–408)).
Bohr, N. (1985). Niels Bohr. In J. Kalckar (Ed.), Collected works (Vol. 6). Amsterdam: North-Holland.
Born, M. (1926a). Quantenmechanik der Stossvorgänge. Zeitschrift für Physik, 38, 803–827.
Born, M. (1926b). Zur Quantenmechanik der Stossvorgänge. Zeitschrift für Physik, 37, 863–867.
Born, M. (1956). The interpretation of quantum mechanics. In Physics in my generation: A selection of papers
(pp. 140–150). London, New York: Pergamon Press.
Born, M., & Heisenberg, W. (1928). La mechanique des quanta. In Électrons et photons. Rapports et discussions du
cinquieme Conseil de Physique, tenu a Buxelles du 24 au 29 Octobre 1927 sous les auspices de l’Institut
Internationale de Physique Solvay (pp. 143–184). Paris: Gauthiers-Villars.
Bromberg, J. (1977). Dirac’s quantum electrodynamics and the wave-particle equivalence. In C. Weiner (Ed.),
History of twentieth century physics (pp. 147–157). New York: Academic Press.
Darrigol, O. (1986). The origin of quantized matter waves. Historical Studies in the Physical and Biological
Sciences, 16, 197–253.
Darwin, C. G. (1927). Free motion in the wave mechanics. Proceedings of the Royal Society, A117, 258–293.
Dirac, P. (1927). The quantum theory of emission and absorption of radiation. Proceedings of the Royal Society,
A114, 243–265.
Dirac, P. (1977). Recollections of an exciting era. In C. Weiner (Ed.), History of twentieth century physics
(pp. 243–265). New York: Academic Press.
Ehrenfest, P. (1927). Bemerkung über die angenäherte Gültigkeit der klassischen Mechanik innerhalb der
Quantenmechanik. Zeitschrift für Physik, 45, 455–457.
Heisenberg, W. (1926a). Mehrkörperproblem und Resonanz in der Quantenmechanik. Zeitschrift für Physik, 38,
411–426. (Reprinted in W. Heisenberg (1985). In W. Blum, H.-P. Dürr, & H. Rechenberg (Eds.), Gesammelte
Werke, Series A/Part I (pp. 456–471). Berlin: Springer.)
Heisenberg, W. (1926b). Quantenmechanik. Naturwissenschaften, 14, 989–994. (Reprinted in W. Heisenberg
(1984). In W. Blum, H.-P. Dürr, & H. Rechenberg (Eds.) Gesammelte Werke, Series B (pp. 52–57). Berlin:
Springer.)
Heisenberg, W. (1929). Die Entwicklung der Quantentheorie 1918–1928. Naturwissenschaften, 17, 490-496.
(Reprinted in W. Heisenberg. (1984). In W. Blum, H.-P. Dürr, & H. Rechenberg (Eds.), Gesammelte Werke,
Series B (pp. 109–115). Berlin: Springer.)
Heisenberg, W. (1930). The physical principles of the quantum theory. New York: Dover.
Heisenberg, W. (1955). The development of the interpretation of quantum theory. In W. Pauli, L. Rosenfeld,
& V. Weisskopf (Eds.), Niels Bohr and the development of physics (pp. 12–29). New York: McGraw-Hill.
Heisenberg, W. (1958a). Atomic physics and causal law. In The physicist’s conception of nature (A.J. Pomerans,
Trans.) (pp. 32–50). London: Hutchinson Scientific & Technical.
Heisenberg, W. (1958b). Physics and philosophy: The revolution in modern science. London: George Allen & Unwin.
Heisenberg, W. (1960). Erinnereungen an der Zeit der Etwicklung der Quantenmechanik. In M. Fierz, &
V. F. Weisskopf (Eds.), Theoretical physics in the twentieth century (pp. 46–47). New York: Interscience
Publishers.
Heisenberg, W. (1983). The physical content of quantum kinematics and mechanics. In J. A. Wheeler, &
W. H. Zurek (Eds.), Quantum theory and measurement (pp. 62–84). Princeton: Princeton University Press
(Originally published as W. Heisenberg (1927). Über den anschaulichen Inhalt der quantentheoretische
Kinematik und Mechanik. Zeitschrift für Physik, 43, 172–198).
Heisenberg, W. (1985). The development of quantum mechanics. In W. Blum, H.-P. Dürr, & H. Rechenberg
(Eds.), Gesammelte Werke Series B (pp. 226–237). Berlin: Springer (Originally published in The Nobel
Foundation (Ed.) (1965). Nobel lectures, including presentation speeches and laureates’ biographies. Physics
1922– 1941 (pp. 290–301). Elsevier: Amsterdam).
Held, C. (1994). The meaning of complementarity. Studies in History and Philosophy of Science, 25, 871–893.
Hund, F. (1974). The history of quantum theory. London: Harrap.
 ARTICLE IN PRESS
K. Camilleri / Studies in History and Philosophy of Modern Physics 37 (2006) 298–315 315

Jammer, M. (1974). The philosophy of quantum mechanics: The interpretations of quantum mechanics in historical
perspective. New York: Wiley.
Jordan, P. (1927a). Über eine neue Begründung der Quantenmechanik. Zeitschrift für Physik, 39, 809–838.
Jordan, P. (1927b). Zur Quantenmechanik der Gasentartung. Zeitschrift für Physik, 44, 473–480.
Jordan, P., & Klein, O. (1927). Zum Mehrkörperproblem der Quantentheorie. Zeitschrift für Physik, 45, 751–765.
Jordan, P., & Wigner, E. (1928). Über das Paulische Äquivalenzverbot. Zeitschrift für Physik, 47, 631–651.
Kennard, E. H. (1927). Zur Quantenmechanik einfacher Bewegungstypen. Zeitschrift für Physik, 44, 326–352.
Kennard, E. H. (1928). On the quantum mechanics of a system of particles. Physical Review, 31, 876–890.
Klein, M. J. (1964). Einstein and the wave–particle duality. The Natural Philosopher, 3, 1–49.
Konno, H. (1978). The historical roots of Born’s probabilistic interpretation. Japanese Studies in the History of
Science, 17, 129–145.
Laue, M. v. (1932). Zu den Erörterungen über Kausalität. Naturwissenschaften, 26, 439–441.
Mehra, J., & Rechenberg, H. (2000). The completion of quantum mechanics 1926–1941: Part I. The historical
development of quantum theory (Vol. 6). New York: Springer.
Mehra, J., & Rechenberg, H. (2001). The completion of quantum mechanics 1926–1941: Part II. The historical
development of quantum theory (Vol. 7). NewYork: Springer.
Murdoch, D. (1987). Niels Bohr’s philosophy of physics. Cambridge: Cambridge University Press.
Pais, A. (1982). Max Born’s statistical interpretation of quantum mechanics. Science, 218, 1193–1198.
Pauli, W. (1979). In K. von Meyenn, A. Herman, & V. F. Weisskopf (Eds.), Wissenschaftlicher Briefwechsel mit
Bohr, Einstein, Heisenberg: 1919– 1929 (Vol. 1). New York: Springer.
Rosenfeld, L. (1979). The wave–particle dilemma. In R. S. Cohen, J. S. Stachel, & L. Rosenfeld (Eds.), Selected
papers of Leon Rosenfeld (pp. 688–703). D Reildel: Dordrecht.
Schrödinger, E. (1928). On the relationship of the quantum mechanics of Heisenberg, Born, and Jordan, and that
of Schrödinger. In Collected papers on wave mechanics (pp. 45–61). London: Blackie & Son Ltd.
Schrödinger, E. (1934). Über die Unanwendbarkeit der Geometrie im Kleinen. Naturwissenschaften, 31, 518–520.
Schrödinger, E. (1957). Indeterminism in physics. In Science, theory and man (pp. 52–80). New York: Dover
Publications.
Schweber, S. (1994). Quantum electrodynamics and the men who made it: Dyson, Feynman, Swinger and Tomonga.
Princeton: Princeton University Press.
Stepansky, B. K. (1997). Ambiguity: Aspects of the wave–particle duality. British Journal for the History of
Science, 30, 375–385.
Wessels, L. (1980). What was Born’s statistical interpretation of quantum mechanics? Philosophy of Science (2),
187–200.