Find Enlightenment

Complementarity: Content, Context and Critique
by Bryan Register

Date: 1 Dec 97
Forum: University of Texas at Austin
Copyright: Bryan Register

Note: The author may or may not still agree with the views expressed in this paper.

Niels Bohr's philosophical notion of 'complementarity' is unusually difficult to understand but quite crucial to our grasp of the history of 20th century physics. In this paper I intend to address four issues: 1. What is complementarity? The philosophical content of Bohr's thought is counterintuitive and his writing is famously bad. 2. What, if any, philosophical influences were operative on Bohr in his development of complementarity? It seems unlikely that such a philosophical approach to a physical problem was developed in a philosophical vacuum. I will consider evidence that Bohr was influenced by Sören Kierkegaard, Poul Martin Møller, William James, and Immanuel Kant. 3. What was the cultural milieu within which complementarity was developed and successfully promulgated? Science does not happen in a cultural vacuum any more than a philosophical one. I will consider the thesis that the central European community of physics was eager for an acausal account of basic reality. 4. Finally, I will consider the larger themes. What does Bohr's thinking tell us about how science works? "What does it all mean?"

I. What is complementarity?

I will offer a characterization of complementarity in my own terms before I turn to the text of Bohr's 1927 Como lecture, in which he lays out the structure of and argument for complementarity. My account is contentious, but I hope that it might add to the clarity and also critique of Bohr's thinking. I wish to discuss Bohr in relation to my own methodological orientation of dialectics . (Sciabarra, Marx, Hayek, and Utopia, 1995)

Dialectics as I will speak of it is not any kind of natural process (as in Hegel or Marx) but rather a methodological research orientation. Dialectical thinking can be loosely characterized as thinking which transcends two dichotomies: the dichotomy between monism and dualism and that between atomism and wholism. A paradigmatic case of dualistic thinking occurs in Descartes's metaphysics, wherein there are two kinds of substance in the world, one kind material and one kind mental. Material substance is that which is extended, mental substance is that which thinks. The two kinds of substance have no means of relating and hence reality is dual in nature. Historically, when a dualism is laid down, later thinkers often develop a reductionist monistic account, collapsing one side of the duality into the other. To continue our philosophical example, Berkeley's metaphysics reduces Descartes' material substance to mental appearances. For Berkeley, all substance is mental. By working a reduction of one of Descartes' dual elements of reality he yields a monistic account. Moreover, there are duels between opposing monisms. For instance, while Hegel regards all reality as a single Spirit, Marx moves from this tradition to the belief that all reality is material. Contrary to these anti-dialectical approaches, dialectics attempts to transcend the duality between dualism and monism themselves. Continuing the example of the material/mental dichotomy, John Searle dialectically regards the mental as irreducible to, yet causally dependent upon, the material. (Searle, The Rediscovery of the Mind, 1992. Searle might not wish to characterize his philosophy of mind as dialectical.) Dialectical thinking attempts to show the causal linkages and lines of dependence between the two sides of an alleged duality, seeking to reconcile the two parts without monistically reducing one to the other.

Moreover, dialectical thinking attempts to transcend the dichotomy between atomistic and wholistic approaches to nature. Atomism regards reality as composed of small parts which are essentially independent from one another. (See perhaps Wittgenstein, Tractatus Logico-Philosophicus proposition 1.21, or Bertrand Russell's Philosophy of Logical Atomism for non-paradigmatic versions of atomism. Atomism rarely, if ever, exists in a pure form.) Wholism regards all of reality as internally interconnected. (See perhaps Brand Blanshard, The Nature of Thought, or much of the Hegelian tradition such as F. H. Bradley.) While atomists regard reality as knowable in little pieces and eschew the need for integration, wholists commit the synoptic fallacy of insisting that the validity of knowledge rests on the possession of all knowledge and sometimes seek an eschatological solution to the problem of knowledge wherein the universe evolves to self-understanding. Dialectical thinking regards reality as interconnected, but maintains that knowledge is possible without omniscience because only some connections are internal; definitive of the thing related. When we consider some piece of reality, we can know it without taking a synoptic or omniscient vantage point by considering it in only its internal connections to the rest of reality, hence not overwhelming our limited mental economy. However, dialectical thinking does not commit the atomist's fallacy of considering a piece of reality in isolation from all else; rather, by stressing the importance of internal connections dialectics allows us to connect the piece of reality under consideration with the rest of reality in essential ways while disregarding inessential factors. (Please note that I am intimately familiar with the distinction between dualists and monists from the study of the early modern period in philosophy. I am trusting Sciabarra's comments for this account of atomism and wholism.)

The essence of complementarity is the view that certain features of the world are 'complementary' in a special sense. I will explain this with the aid of the wave-particle nature of light. When we examine a beam of light in ways appropriate to measuring a stream of particles, we obtain certain results (such as particular impacts on the measuring apparatus) consistent with the hypothesis that light is a stream of particles. When we measure a beam of light in ways consistent with the hypothesis that light is a wave, we obtain results (such as interference patterns) consistent with the hypothesis that light is a wave. There is no means available to us to measure light in both its wave and particle aspects. For complementarity, these are two complementary approaches to describing reality. However, there can be no progress beyond this irreducible duality. The description of reality requires explanation in terms of either waves or particles, because these are concepts developed in our ordinary experience of the macroscopic world. A new conceptual apparatus which treats light as neither a wave nor a particle, but as a third, indescribable, phenomenon which acts sometimes like a wave and sometimes like a particle, is ruled out on grounds of incoherence with our ordinary experience of the world in terms of which all experimental results must be reported. Thus, complementarity rightly rejects a monistic reduction on grounds of empirical evidence, but refuses to accept a non-intuitive or purely mathematical description of quantum reality because of considerations of the meaning and origin of concepts. (I return to these considerations in more detail on pp. 15-17.) Complentarity thus provides a dualistic account of quantum reality.

Moreover, classical mechanics deals with light as waves, while it is the new quantum mechanics which requires a description of light as particles in addition to waves. But the behavior of light in quantum reality is not consistent with classical mechanics except in the limit wherein quantum reality approaches the classical description. If this is the case, then quantum reality cannot be understood simultaneously with an understanding of classical reality. This corresponds to the distinction between atomism and wholism. To whatever degree we are investigating reality in a way which coheres with the particle nature of light, we can know quantum reality but we lose its connection to the rest of reality. However, when we investigate reality in a way which coheres with the wave nature of light, we can know classical reality but lose the details of the quantum situation. Since the modes of investigation are complementary and hence irreconcilable, and the atom is provided by the one and the whole provided by the other, there is no way to dialectically consider the part in its essential context to the whole. Hence I conclude that complementarity is exactly opposed to a dialectical mode of investigation, which might, with Duhem (Aim and Structure of Physical Theory , 1905) advocate a purely mathematical account with no requirement of visualizability. (See also pp. 15-17.) (David Bohm's account may be dialectical; it might be instead a wholist account, see The Undivided Universe .)

Before continuing with Bohr's argument, I must digress to make several points drawn from quantum mechanics. These points are non-philosophical in that they are observation reports not yet interpreted in any relevant way.

Historically, there had long been a debate between proponents of light as a wave and light as a beam of particles. The classical physics of the 19th century had regarded light as wave-like. New experimental reports in the early 20th century indicated that light had a particle-like nature. To understand Bohr's work, we must glance at both of these characterizations and their relationship.

In the wave characterization of light, its frequency (as that of all waves) varies directly with energy: high energy light has high frequency. Frequency is inversely related to wavelength: high frequency light has low wavelength. A sort of bridge between the wave and particle characterizations is that wavelength varies inversely with the 'compactness' of a photon, or particle of light. Thus, higher-energy photons are more compact, lower-energy less compact. If light is regarded as a wave, high-energy waves are short but high, while low-energy ones are low but long. If light is regarded as a particle, high-energy photons are small and dense, while low-energy ones are large clouds. Light can be regarded as either waves or particles, but not at the same time. For Bohr, then, these are complementary descriptions which are irreducibly dualistic.

Particles at the quantum level are intriguing in that they are not observable by any passive means. While even living cells can be observed with the aid of a passive microscope which does not interact with the cell, sub-atomic particles can only be studied by active modes of investigation. To measure features of an electron such as its position and momentum, we must force it into interaction with photons. This is something like radar or sonar in which we send out a sensing beam which carries information back to us after having interacted with objects (in radar and sonar, by bouncing back from them). The unique problem in quantum measurement is that the photons with which electrons must interact to be measured are on the same order of causal power as the electron itself. When the photon and electron interact, we benefit from the interaction in that the electron has affected the photon in a way informative to us. But since we are using such a gross measuring technique, the electron has been affected by the photon with which it interacted. The reason radar does not knock down the planes whose position it discovers is that the radar waves are very weak in comparison with the planes. If we had to find something's position by, say, waving a baseball bat about until it smashed into something, this would be analogous to the effect of the measuring photon on the measured electron.

The measurement problem deepens when we consider another factor. To see how electrons work, we want to know (at least) two things about them: their momenta and their positions. We can find these by forcing electrons into interaction with photons. But, unfortunately, we can determine only one of momentum or position at a time. As one might expect, when using a compact, high-energy photon to measure the momentum or position of an electron, the photon has a high momentum and hence measures only the position of the target electron clearly, but cannot precisely determine the momentum of the target electron. Using a dispersed, low-energy photon pinpoints momentum with great clarity but cannot determine the electron's position. The degree to which we measure position is the degree to which we do not measure momentum, and conversely. The relationship here is one of uncertainty: to whatever degree we know the precise momentum, we do not know the precise position and contrariwise. Since Werner Heisenberg discovered this crucial relationship it is called the "Heisenberg Uncertainty Principle".

There is one last problem with quantum measurement. Not only do photons come with certain inverse relationships, they only come in certain allowable levels of energy. In an atom, electrons can exist only at certain energy levels, called orbitals. The first orbital has members of only a certain very low energy; the second has members of only a certain somewhat higher enery, and so forth. Electrons move from higher levels to lower by emitting a photon. They move from lower levels to higher when absorbing a photon. Photons are packets of energy, and so the amount of energy emitted or absorbed by an electron in moving between orbitals must match the amount of energy required to move to the lower or higher orbital. Thus photons exist only in the quantities of energy required to raise or lower an electron by one orbital. It is these allowable energy levels which Bohr calls the 'quantum of action'. Not only can we learn position or momentum of electrons only to the degree that we have not learned the other, we can learn them only within certain levels of precision.

In introducing his work, Bohr tells that the Como lecture, his first presentation of complementarity, was conceived in opposition to Schrödinger's wave mechanics. In the mid-twenties crisis of physics,

Especially had the great success of Schrödinger's wave mechanics revived the hopes of many physicists of being able to describe atomic phenomena along lines similar to those of classical physical theories without introducing 'irrationalities' of the kind which had thus far been characteristic of the quantum theory. In opposition to this view, it is maintained in [the Como lecture] that the fundamental postulate of the indivisibility of the quantum of action is itself, from the classical point of view, an irrational element which inevitably requires us to forego a causal description and which, because of the coupling between phenomena and their observation, forces us to adopt a new mode of description designated as complementary in the sense that any given application of classical concepts precludes the simultaneous use of other classical concepts which in a different connection are equally necessary for the elucidation of the phenomena. (Atomic Physics and the Description of Nature, 1934, introduction, p. 10)

Bohr reports that, in the historical context, Schrödinger's work had provided an alternative to Heisenberg's matrix mechanics. Many physicists believed that Schrödinger's work meant that quantum mechanics would not require a radical break with classical mechanics. Since classical mechanics had regarded light as waves, regarding the new quantum phenomena as irreducibly wave-like would not require the violation of classical modes of thought. It is this position which Bohr takes as his foil.

Bohr maintains that the quantum of action, or the fact that electrons within atoms jump from orbital level to orbital level without occupying intervening spaces or taking time for the jump, implies that there can be no causal account of electron motion. This is because for him causality is an impoverished notion of mere direct contact pressures. A causal account for Bohr is an account of the motions of things as a result of their interaction with other things. The paradigm case of causal interactions is the interaction of billiard balls: one object hits another, makes it move, and itself stops moving. But in quantum measurement, we must choose between location and momentum, which latter includes motion. The precision of our knowledge of the one is the imprecision of our knowledge of the other. If we find location, we can't know motion and hence can't discover causal relations between particles. To repeat the billiard-ball analogy, we can get location by taking only a still photo of the table, which tells us nothing about where - or even whether - the balls are moving. An experience of momentum, on the other hand, would be nothing but gigantic vectors drawn on a schematic drawing of the table, letting us know which direction the balls are moving but not where they are. Neither of these accounts are causal or allow for a causal interpretation (even in Bohr's sense, because knowing that there is causal contact rests on knowing that there is contact, which is a matter of location), but for Bohr it seems that the momentum description is a causal account despite the absence in it of a description of location. In these comments, Bohr seems to be freezing the concept of causality into one of its instantiations; the version of causality which appears in Cartesian mechanics and which maintains that causal relations are exclusively contact pressures between physical objects and fields. It is worth noting that, on this implicit account of causality, Newtonian gravitation is a non-causal relation because it operates instantaneously and at a distance. A richer concept of causality, such as the Aristotelian (in which the cause of action is not exclusively prior action but the nature of the acting entity), might enable a causal account of electron motion.

Bohr argues that there is a difficulty decoupling the observer from the observed. He explains that his ground for this is that "the finite magnitude of the quantum of action prevents altogether a sharp distinction between a phenomenon and the agency by which it is observed " (Atomic Physics and the Description of Nature, introduction, p 11). Since electrons and the photons by which we measure their momenta and positions exist only in certain allowable energy levels, there is a certain acontinuity built into the structure of matter at the atomic level. Bohr's point is that, since in all quantum measurement is active, the activity undertaken by the subject (scientist) in attempting to apprehend the object determines the object (electron). Hence, the object is a product both of its own intrinsic nature and the active means of knowing it employed by the subject, making it difficult to distinguish the object from the subject.

There are two critiques which can be raised against this argument. The first is that it is not the quantum of action but rather the inverse relation between the measuring capacities of photons which yields Bohr's epistemological problem. Because of the quantum of action, there are specific allowable levels of energy of photons. Let us say for simplicity's sake that these allowable levels are whole numbers: 1, 2, 3, 4. Since high-energy photons measure position and low-energy photons measure momentum, photons with energy level 4 would measure position but not momentum and photons with energy level 1 would measure momentum but not position. This relationship between 1 and 4 would hold even if non-whole numbers were allowable levels of energy; that is, if photons could exist with any quantity of energy and there were no quantum of action. The relationship between 4 and 1 is not changed by the possible existence of 1.5. Moreover, we could use a photon of energy 2 to discover fairly detailed information about position but only vague information about momentum, or we could use a photon of energy 3 to discover fairly detailed information about momentum but only vague information about position. But we could not use a photon of energy 2.5 to discover an in-between level of information. The precision of our measurement is determined by the energy of the photon used. Thus, since photons come in only certain allowable levels of energy, only measurements of certain precisions are possible. Bohr believes that this feature, the quantum of action, is of crucial epistemic importance. But the quantum of action merely stipulates a limit on the precision of our measurement and cannot justify Bohr's subject/object difficulties. That an object can be known to us only so precisely does not mean that it is indistinguishable from the subject that knows it. The deeper structure of argument is that by choosing which level of energy to employ in measuring the electron, we choose what to measure. It is this relationship which is epistemically relevant, not the fact that photons only come in certain allowable energies. But even this relationship is unexciting, as it only stipulates a limit on which properties of the object are knowable. This does not demand a radical reconsideration of the subject/object relationship: that we can know only some things about an object does not mean that it is not an object distinct from the subject which knows it.

Secondly, Bohr is wrong to argue that since we must actively interact with the object to measure it, the object is partly a product of the subject's activity and hence is partly subjective. That his argument deals in particular with light interacting with the object tips us off to the problem. We are only able to see macroscopic objects when they are exposed to light, as well. Since it is is in virtue of these conditions that the thing becomes an object, these conditions cannot be regarded as invalidating the objectivity for which they are the prerequisite. Traditionally, accounts of objectivity have demanded that an object be given to the subject without in any way interacting with the subject; a clearly impossible state of affairs. A more novel account of objectivity (such as Kelley's in The Evidence of the Senses ) might allow us to regard the object as directly given despite its causal interaction with the subject. On such an account, there is no tension between describing the thing as it is in itself and the thing as it is given us, because we can describe the thing as it is in itself because it is directly given us through our cognitive faculties in a causal interaction. There is no principled distinction between macroscopic phenomena which become objects to us without complementary descriptions but only when light is present, and the measurement of quantum phenomena. Quantum phenomena can become an object to us only under certain conditions, just as classical phenomena become objects to us only under certain conditions. Those conditions cannot therefore invalidate objectivity but are objectivity's prerequisite. Complementarity as a research program holds that we are limited to alternating wave and particle descriptions of quantum reality. But it may be possible to maintain a fully consistent description of quantum phenomena which involves neither wave nor particle descriptions. Such a description would not employ complementary concepts such as wave and particle, but would attempt to dialectically relate all known features of quantum phenomena within a new (possibly non-intuitive or exclusively mathematical) conceptual apparatus. If this program could be successful, then complementarity hinders the investigation of quantum reality.

I have now completed a critical and speculative review of the content of complementarity. A more full statement, in Bohr's own words, can be found in the first part of his Como lecture, pp. 52-57 of Atomic Physics and the Description of Nature . According to Folse (The Philosophy of Niels Bohr, 1985) "Bohr presents virtually the entire argument [of complementarity] in the six paragraphs which comprise the first section of his paper." (p. 108) Moreover, Folse notes that this notion of complementarity was "never seriously altered" (p. 105) and that "The argument presented in the Como paper always remained Bohr's basic approach to complementarity." (p. 118) Folse provides a clear explication of this crucial passage (pp. 108-117) on which I based the above discussion of complementarity. I hope that my critical comments have made clear that Bohr's notion of complementarity cannot be simply read from the conditions of quantum measurement. Since this is the case, I will turn now to consider possible influences on Bohr's thinking from outside the physical problems. It is possible that some outside force informed Bohr's interpretation. In any case, the figures of whom I will speak may have influenced Bohr even if I have made serious errors in my interpretation of complementarity. However, I maintain my interpretation throughout and the reader should note that my cases for influences will be weakened in important ways if my interpretation is in error.

II. Whence Complementarity?

Any hypothesis about philosophical influences on Bohr is inherently speculative. Bohr did not leave anything like detailed notes on the works of earlier thinkers and his colleagues do not report anything like Bohr reading scholarly commentaries of Kant's critiques or other philosophical works. Moreover, his interviewers have found only hints of evidence for influence, since Bohr passed away before the interviewing process was complete.

Any influence is very likely to have worked on Bohr through the medium of his philosophy professor and family friend Harald Høffding. Pais (Niels Bohrs' Times, 1991) notes that Bohr took Høffding's required freshman philosophy course but had already known him:

Bohr had known... Høffding long before becoming a university student. Christian Bohr [Niels's father], Christiansen [Bohr's physics professor], Høffding, and the famous linguist Vilhelm Thomsen would gather at the home of one or the other for further discussion... Høffding apparently appreciated young Niels, for at one time during the latter's student years he sent him some sheets of a new edition of his book on logic with a request for his 'customary criticism'. Two months later he wrote Bohr thanking him for his 'good collaboration'. In later years Bohr spoke with great respect of Høffding, admiring his ever searching open mind, in particular his efforts at understanding the principles of quantum mechanics. He once took Heisenberg along to listen to a 'beautiful lecture on Socrates' by the 85-year-old Høffding. When Bohr visited Høffding during his final illness, he read poetry to him. (p 99)

Clearly, there was an intimate relationship between the two men and it would be hopelessly strange to maintain that the two did not discuss philosophical issues. Moreover, there is scant reason to suspect that Bohr had any scholarly contact with philosophy save through Høffding. We should therefore turn to Høffding's work, most of which has not been translated, for some idea of who might have been at work on Bohr's mind. Bohr's association with Høffding suggests three influences: Søren Kierkegaard, William James, and Immanuel Kant. Poul Martin Møller is another influence, but he may not have come to Bohr through Høffding.

Gerald Holton discusses the possibility of a Kierkegaardian influence on Bohr in his paper "The Roots of Complementarity." (Dædelus 99, Fall 1970) Holton notes that, "In Høffding's own life, a crucial and early influence was the work of Kierkegaard... Høffding reported that in a youthful crisis, in which he was near 'despair,' he had found solace and new strength though Kierkegaard's writings..." (p. 1040) Moreover, we can be absolutely confident that Kierkegaard impressed Bohr as well. As Holton points out, " is remembered that in 1909 Niels sent his brother Harald as a birthday gift Kierkegaard's book, Stages on Life's Way, with a letter saying 'It is the only thing I have to send; but I do not believe that it would be very easy to find anything better.... I even think it is one of the most delightful things I have ever read.'" (pp. 1040-1041) Pais demurs, arguing that Bohr said that he was "not in agreement with Kierkegaard". But Pais's other quotation is different: Bohr also says of Kierkegaard that "His language is wonderful, often sublime [but] there is of course much in Kierkegaard which I cannot accept." (Pais, 424) I would argue with Holton that "One can well imagine that Niels Bohr could enjoy the aesthetic experience and the moral passion, without having also to agree with the antiscientific attitude of much of the work." (Holton, p. 1041) Further, one can disagree with the content of a theory while accepting its structure. Kierkegaard maintained a stage theory in which individuals can progress in life through a series of incommensurable stages; we leap from stage to stage in analogy to an electron leaping from orbital to orbital. In this light it is trivial to see that reading Kierkegaard only a few years before he developed his 1912 quantized theory of the atom may have been an influence on Bohr's thinking. (For accounts of Kierkegaard's thought, see David Norton, Personal Destinies, 1976, pp. 64-79, and especially Høffding, History of Modern Philosophy, volume 2, 1900, pp. 285-289.)

I want also to mention a much more speculative hypothesis regarding Kierkegaard's influence; one which is ill-supported but intriguing. Recall that, for Bohr, a causal description of electron motion in quantum reality is impossible. This implies that something other than mechanism is involved in moving an electron about. Moreover, Kierkegaard's romantic existentialism maintains that an individual's change from one stage of life to another is a matter irreducibly of subjective choice. This mirrors Bohr's notion of irreducible acausality in electron motion. Bohr notes in the Como lecture that

Here the contrast with the ordinary way of description appears strikingly in the circumstance that spectral lines, which on the classical view would be ascribed to the same state of the atom, will, according to the quantum postulate, correspond to separate transition processes, between which the excited atom has a choice ." (Atomic Theory and the Description of Nature, p. 69, emphasis added)

This is probably a metaphorical description, but at least in Bohr's linguistic choice it seems that electrons move freely and without cause between incommensurable stages just as does the human soul in Kierkegaard. The hypothesis that Bohr developed the incommensurability of quantum orbitals from Kierkegaard's account seems likely. The hypothesis that he developed the acausal nature of the shift under Kierkegaard's influence seems on face less likely but a consideration worth taking note of.

I want to mention now the, as Pais puts it, "One Dane whose influence Bohr acknowledged... Poul Martin Møller, an early nineteenth century poet and novelist, and at one time professor of philosophy at Copenhagen, whom [Bohr] would often quote." (424) According to Pais, the oft-quoted lines are these from Møller's novel Adventures of a Danish Student :

My endless inquiries made it impossible for me to achieve anything. Moreover, I get to think about my own thoughts of the situation in which I find myself. I even think that I think of it, and divide myself into an infinite retrogressive sequence of 'I's who consider each other. I do not know at which 'I' to stop as the actual, and as soon as I stop, there is indeed again an 'I' which stops at it. I become confused and feel giddy as if I were looking down into a bottomless abyss, and my ponderings result finally in a terrible headache. (Pais, 440)

Møller is confronted with the 'problem of the disappearing self'. This is the problem of trying to experience oneself. I will explain in steps:

One can continue this process ad infinitum but one will never be able to observe the awareness with which one must be identified in the moment. Subjectivity is never an object. Hence's one's self of the moment has always disappeared from one's gaze just as it forms, because it becomes the gaze itself. Trying to make an object of one's own subjectivity is futile: one might as well try to see what one looks like with one's eyes closed by standing in front of a mirror and opening one's eyes very quickly. As soon as one's eyes open, the object (one's face with closed eyes) is ruined; analogously, as soon as one has found subjectivity, it has become an object.

It is intriguing that Bohr would cite this passage, because he seems to be faced with exactly the opposite problem, the problem of a disappearing object. When we look at quantum phenomena, we find as much evidence of our own measuring apparatus (which seems implicitly to be for Bohr an extension of the subject) as we do of the object as it is in itself. Where, then, is the object? For Bohr, the object of one's awareness is in part the subject himself. But for Møller, the subject can never be an object because the subject disappears under its own gaze. Combining these views brings an endless circle of self-referential negation: the subject, when it gazes at some feature of itself, reduces that feature to a mere object; an external thing taken as an object becomes, in virtue of being taken as an object, a feature of the subject, but when the subject gazes at that feature of itself that feature becomes thereby a mere object, and so forth. The subject annihilates itself in favor of being an object, but an object is always annihilated in favor of being a feature of the subject, which subject anihilates itself.... According to Pais, Bohr quoted this passage frequently, so we are left no choice but to accept that Bohr appreciated it. I will try to show the full implication of this when concluding our discussion of James.

We need first to know when Bohr read James. The following is from the final interview of Bohr, conducted by Thomas Kuhn:

Bohr: I was a close friend of Rubin [a fellow student, later psychologist], and, therefore, I read actually the work of William James....

Kuhn: When was this that you read William James?

Bohr: That may be a little later, I don't know. I got so much to do, and it may be at the time I was working with surface tension [1905], or it may be just a little later. I don't know.

Kuhn: But it would be before Manchester [1912]?

Bohr: Oh yes, it was many years before. (pp. 1034-1035)

So it seems that Bohr read James fifteen years or more prior to the 1927 Como lecture. However, Holton points out that:

Rosenfeld [in a letter to Holton] has expressed his strong belief that the work of William James was not known to Niels Bohr until about 1932. He recalls that in or about 1932, Bohr showed Rosenfeld a copy of James's Principles of Psychology . Rosenfeld believes that a few days earlier Bohr had had a conversation with Rubin, the psychologist and Bohr's former fellow student. Rubin may have sent the book to Bohr after their conversation. Bohr showed excited interest in the book, and especially pointed out to Rosenfeld the passages on the "stream of consciousness." During the next few days, Bohr shared the same excitement with several visitors, and Rosenfeld retained the definite impression that this was Bohr's first acquaintance with William James's work. (p. 1035)

But 1932 is five years after the Como lecture and also several years after Bohr had begun to apply complementarity to psychological issues. (Pais, 439-441) Are we to believe Bohr or Rosenfeld?

As Folse notes, however, Høffding had visited James in America in 1904 and had a very positive view of his work. Moreover,

That Høffding would have then returned to Denmark and made no mention of James to his circle of students and friends, including the Bohrs, both father and son, seems completely incredible.... It seems extraordinary to suppose that both Rubin and Høffding would have listened to Bohr's meditations on these themes and not noticed the obvious connection with James's problems in The Principles of Psychology. (50-51)

Folse hypothesizes that Bohr may have been aware of James's psychological work as early as 1904, but may not have discovered James's philosophical work until 1932. According to Folse, this accords nicely with both Rosenfeld's and Bohr's memories, because Bohr speaks of James's psychological work whereas "when Rosenfeld brings up James's name, he does so in connection with the claim that complementarity is essentially a pragmatist character..." (51) However, Folse seems to be in error. Bohr recalled having looked at "The Stream of Thoughts" before 1912, while Rosenfeld believes that Bohr first looked at The Principles of Psychology no earlier than 1932. But "The Stream of Thoughts" is a chapter of The Principles of Psychology and hence their memories do in fact conflict.

But Rosenfeld's memory seems to be much more distinct than Bohr's. In the interview with Kuhn, Bohr talks like this: "I read some, but that was an interest by [and here Bohr suddenly stopped and exclaimed]-oh, the whole thing is coming [back to me]! ...I think I read the book, or a paragraph, called... No, what is that called? It is called 'The Stream of Thoughts'..." (Holton, 1034) Here we seem to have exactly equal evidence for the two sides. On the one hand, Folse argues that it is improbable in the extreme that Høffding would not speak of James to Bohr very early on. On the other hand, Rosenfeld's recollection seems more reliable than Bohr's.

I am going to move now to a consideration of what content Bohr's thought might have were he influenced by James at an early date. We must remain agnostic about the question of influence, but it is likely that Bohr had had at least some exposure to James's thought early on, even though his full examination of James's work seems likely to have occurred too late for an influence to have occurred. Moreover, it is at least slightly possible that Bohr discovered James twice, once in the 1900's and again in 1932. If his early exposure was not deep or rigorous, he may have simply forgotten the text in the intervening quarter of a century.

Holton cites from James: "Consciousness does not appear to itself chopped up in bits; it flows. Let us call it the stream of thought, of consciousness, or of subjective life." But in apparent contrast, James also wrote that "Like a bird's life, [thought] seems to be made of an alternation of flights and perchings. The rhythm of language expresses this, where every thought is expressed in a sentence and every sentence closed by a period... Let us call the resting places the 'substantive parts,' and the places of flight the 'transitive parts,' of the stream of thought." (Holton, 1036) Which is it? On the one hand, consciousness flows in an unbroken and integrated stream much like the wave-oriented classical description of the world, on the other, consciousness works in fits and starts rather like the particle-oriented quantum description of the world. This is a perfect match to the irreducible duality demanded by complementarity, providing additional evidence for some early influence.

Moreover, according to Holton, James moves into a problem identical to Møller's disappearing self. To try to catch the transitive parts is to make them all of a sudden into substantive parts. Hence the transitive parts of consciousness - which presumably include actions like perception, deduction, introspection and so forth - always shift out from under one's gaze to become the gaze itself. We can have no experience of the active knowing self.

Møller and James both experience the problem of the disappearing self. It is very difficult to say in what way Bohr's problem is related. On the one hand, the problems are exactly opposite: Bohr can't find an object, the others can't find the subject. On the other hand, the problems are structurally identical: Bohr's objects disappear under his gaze while the others' own subjectivities disappear under their own gaze. I will now combine the two views to try to see a coherent influence; to try to maintain that Bohr accepted both theses. Since the object of quantum measurement melts into the subject at the same time that the subject turns into an object when examined, the object of quantum measurement is simultaneously an object which is presented by consciousness to itself, and intrinsic to the subject. That is, all experience is of the subject, which ceases to be subjective when examined and is hence an object. This is an endless and maddening circle.

Let me try to clarify this claim. For Møller and James, the self disappears under examination but the object is a steady reference point. We may disappear in an infinity of introspection, but there is always the object to which we inevitably return. For Bohr, however, the object just is a feature of the subject. If Møller and James provide us with a metaphysics of subjective vertigo, Bohr provides us with a metaphysics of pure absence: endless falling with no chance of ever hitting the ground of the object - and with no substantial self to have the chance.

Let me conclude the question of influences with a discussion of Bohr's relation to Immanuel Kant. I will argue that Bohr was operating under implicitly Kantian modes of thought and that Bohr's philosophy was Kantian through and through both in method and content. I see little reason, however, to maintain that he was directly influenced by Kant as he may have been by the previous three figures.

Honner claims in The Description of Nature (1987) that Bohr's argument for complementarity is a transcendental argument which takes the same form as Kantian arguments for synthetic a priori knowledge. I will try to show a methodological similarity between Kant and Bohr by discussing Honner's gloss of Bohr's argument and juxtaposing it with key ideas from Kant's work. My argument for similarity in content will stem from my discussion of James and Møller's contribution above.

Honner comments that:

Kant and Bohr are both concerned with what cannot but be the case. For Kant this necessity is not an empirical demand, though it is connected with the necessary conditions for the possibility of experiential knowledge. For Bohr the necessity is both empirical, inasmuch as it rests on the quantum postulate, and epistemological, given his elaboration of the necessary conditions for unambiguous communication about experience. Kant gave the term 'transcendental' a specific, albeit elusive, meaning. He made a sharp distinction between that which lies entirely outside the sensible world, which he called 'transcendent', and that which in our reasoning refers to the 'totality of conditions in the sensible world', the 'transcendental'. (p 12)

For Kant, sense experience is generated by the subject under some kind of stimulus from the thing-in-itself, which is always outside of experience. The thing-in-itself is transcendent. However, certain considerations which are not transcendent are implied in all reasoning and all experience; that is, they transcend the known without going beyond it. These considerations are known as the transcendental. It is the transcendent thing-in-itself which makes it such that we have experience at all; it is the transcendental forms of the mind which determine what that experience will be.

Moreover, Honner points out that "...Kant turned to his transcendental approach. He suggested that we begin with the givenness of experience and then, by reflecting on the necessary conditions for the possibility of experience, move to universal conditions." (p. 12) Kant sought conclusions which would be necessary and yet possessed of empirical significance. Historically, empiricists had denied the possibility of necessary truths and rationalists who sought necessary truths did so by stipulating some basic definitions without empirical ground and proving theses based on them. The one was critical, the other dogmatic, and they chased one another like a dog his tail: in a very small circle. Kant sought to break out not by trying to prove necessary truths from individual experiences (assumed in advance to be impossible) but by reflecting on that which makes our experience possible. He sought not the conclusions of experience but the presuppositions of experience. Such truths would be necessary for us, since our experience would always adhere to them, but would possess empirical import, because they state that to which experience must adhere.

According to Honner, Bohr "constantly argues through a series of ten steps" (p. 104) but Honner notes that "After recognizing these links, it is possible to arrange Bohr's arguments more economically, so that the connections between key statements become more apparent." For simplicity's sake, I'll just move through the shorter three-term version and compare it with elements of Kant's program.

(B1) Some kind of conceptual framework is a necessary condition of the possibility of ordering experience.

It is possible that perceptually given phenomena are ordered, not by a pre-existing conceptual apparatus as in Kant, but by their own ordered nature and the capacity of the mind to recognize this pre-existing order. But with more complex phenomena, causal and other relations can be understood only with the aid of a conceptual framework. It is impossible to perceive, for instance, the effect of a certain monetary policy on the economic situation. But Bohr is attempting to find a method for describing very complex problems, the interpretation of which requires a very complex and delicate conceptual apparatus, so the claim clearly holds true for his problem.

(B2) It is a necessary condition of the possibility of objective descriptions of processes at the boundaries of human experience that concepts related to more normal experiences be employed.

This gloss is more difficult to interpret. What exactly are the 'boundaries of human experience'? Honner's Bohr contrasts them with 'more normal experiences', seeming to imply a temporal relationship: the normal is that with which we already have experience (such as classical mechanical phenomena), the boundaries are the new frontier experiences (such as quantum mechanical phenomena). Bohr is arguing, then, that new experiences such as experience of quantum phenomena must be interpreted in terms derived from classical mechanics if the description is to be objective, that is, if the interpretation is to make the quantum event into an object of our awareness. This is a Kantian claim wherein the possibility of something's being an object of our awareness rests on the object matching (in some way) the nature of the subject, not the other way 'round. The 'nature of the subject', in this case, is the pre-existing conceptual apparatus.

(B3) Our position as observers in a domain of experience where unambiguous application of concepts depends essentially on the conditions of observation demands the use of complementary descriptions if the description is to be exhaustive. (p 105)

This reiterates the deeper structure of argument which I found in complementarity, above on pp. 7-8. The 'conditions of observation' to which Honner refers are the measuring apparatus which relate inversely and measure different features of quantum events. Since the opposed conditions (choices of measuring apparatus) provide us with different information, an exhaustive description must combine the descriptions provided by both of the possible conditions. In this case, the two possible conditions tell us that light is either a wave or else it is a particle, hence we must describe light under each aspect to complete its description.

The key step in the argument is the claim that objectivity of description requires conservatism in conceptual use. Honner had previously interpreted this point as point (vi) in the longer version of his gloss, which reads:

(vi) It is a necessary condition for the possibility of unambiguous communication that suitably refined everyday concepts be used, no matter how far the processes concerned transcend the range of ordinary experience. (p. 88)

Honner further notes that

The argument is not an axiomatic one, but a direct articulation of what he sees as constitutive of experiment and communication.... He takes it as read that ordinary language can be used to provide a satisfactorily objective account of experience and activity in the real world. Further, he appears to assume that any successful communication must employ terms which are related to such ordinary language. (p. 88)

The last sentence is Bohr's Kantian claim. Objectivity for Kant rests on an experience being informed by the categories of the mind, while for Bohr objectivity rests on an experience being informed by the categories of classical mechanics. The resemblance may be even more thorough-going when we bear in mind that Kant's categories were developed with a Newtonian or classical universe in mind.

On my interpretation of Kant's Critique of Pure Reason, such elements as the Transcendental Aesthetic and Transcendental Deduction(s) do not function as arguments in the sense of deductive motions from premises to conclusions or gatherings of empirical evidence, but rather a meditative reflection on human experience to see ever-present within it certain structural conditions, the categories of the mind. Bohr seems to make a similar claim. When we reflect on the conditions of experimental reports, we always find the classical phenomena of wave and particle, but quantum phenomena evades description by either of these concepts alone. Hence, the possibility of a quantum phenomenon becoming an object to our awareness - becoming objective - is our realization of it in classical terms treated as complementary to each other. This is a highly Kantian argument and completes my justification for the conclusion that Bohr was a methodological Kantian.

But the method and the content of complementarity are not the same thing and I will turn now to discuss one last time the content of complementarity. Method and content exist in a dialectical relationship such that each tends to determine the other. What content must Bohr's philosophy have, if its method is as I have described above?

In quantum phenomena, what it is that becomes an object to the subject depends on the action of the subject. Does the subject choose to experience a wave or a particle? Position or momentum? The thing-in-itself is determined as one of these not solely in virtue of its intrinsic nature but in virtue of the knowing activity of the subject. It is difficult to find in the quantum object something not placed in it by its examination by the subject. This is why the subject and object are tightly coupled: the object is just as much a feature of the subject as it is of the intrinsic thing-in-itself. This is the deep subjectivity which I described on pp. 14-15 in reference to James and Møller.

Likewise for Kant. While Kant's categories of the mind are not mutually exclusive in one experience, they determine what we experience. The object of one's awareness is one's own subjectivity. Moreover, as on pp. 14-15, the knowing subject is ultimately unknowable but is itself a thing-in-itself our knowledge of which is determined by the categories. Kant's philosophy is as vertiginous as I tried to show Bohr's to be above. I take this as a proof of content similarity.

The main difference between Kant and Bohr comes specifically on the issue of whether the determinants of experience are complementary or mutually consistent. For Bohr, space-time location and causality are complementary. Recall that location and momentum are complementary and mutually exclusive descriptions of electrons. For Bohr, it seemed that a momentum description was to be the causal description. So location in space and time could not be known along with the causal relations acting on an electron. But for Kant, space and time inform every external experience and causality is the rule according to which phenomena are placed in space and time. Space and time alone are insufficient to create our experience; by themselves, they would allow for my dropped glass to turn into a rhinoceros on its way to the ground, levitate for twenty years, fly back up to my hand, and shatter there without ever having hit the ground (or turned back into a glass). Kantian causality is the mental function according to which phenomena are ordered coherently in space and time, according to rules which probably amount to classical mechanics. This seemingly crucial point is actually insignificant, as it is unlikely that Bohr studied Kant in sufficient depth to even know what his account of causality was.

This brings us to the question of how Bohr may have been exposed to Kant. Recall that Bohr took Høffding's course in the history of philosophy. Høffding's main translated work is his two-volume History of Modern Philosophy, volume two of which includes almost a hundred pages devoted to Kant. He says of Kant that "...the point of view which he adopted possesses general significance for the study of religion, art, literature, and languages, as well as for the study of institutions and the forms of society." Moreover, Kant's central theses about the workings of the subject in producing experience "lay at the heart of Kant's great and profound intellectual work, and it is in virtue of this that he occupies the central position in the history of modern thought..." (p. 31) While Høffding is hardly a Kantian, we may be sure that Bohr gained at least some access to Kant's central theses in his university class.

I have argued for possible influence on Bohr by Kierkegaard, Møller, James, and Kant. My arguments have been that Bohr's work showed a certain similarity to the work of the prior thinker, and that there is historical evidence for Bohr having studied the thinker's work. We cannot ignore, of course, the problems in quantum mechanics which stimulated Bohr's meditations. I do not mean to suggest that Bohr did nothing but produce a version of others' thoughts; rather, he responded to problems in the way one might were one influenced by these figures.

III. Why Was Complementarity Accepted?

Bohr's ideas were ultimately accepted by the physics community as a whole, with the exception of Einstein and a few others. We have seen that complementarity cannot be read out of quantum phenomena but that there are other factors which gave rise to it. We cannot believe, however, that every physicist but Einstein had been subject to the same constellation of influences which Bohr had been when he developed his views. What explains the fairly ready acceptance by the physics community of this perspective which is so counterintuitive?

I will argue with Forman ("Weimar Culture, Causality, and Quantum Theory, 1918-1927: Adaptation by German Physicists and Mathematicians to a Hostile Intellectual Environment", Historical Studies in the Physical Sciences, 1971) that following the defeat of Germany in the First World War the German academic establishment grew hostile to the idea of causality and sought a romantic, as opposed to scientific, explanation of the world. This was then the intellectual milieu within which quantum mechanics was discovered and its interpretation promulgated.

Forman tries to show that the aftermath of Germany's defeat the dominant intellectual tendency in the Weimar academic world was a neo-romantic, existentialist "philosophy of life," reveling in crises and characterized by antagonism toward analytical rationality generally and toward the exact sciences and their technical applications particularly. (p. 4)

During the war, hard scientists had stressed their contribution to the war and the impending victory as a victory of science and technical rationality. When the war came to its famously disastrous conclusion, science, having sought the credit, took some of the blame. Forman notes that scientists were blamed for "the destruction of the soul, of the world" and that one scientist noted at the time that opponents of science treated it " bearing the guilt for the world crisis in which we stand at present, and the whole of the intellectual and material misery bound up with that of the crisis is charged to natural science's account." (p. 11)

A paradigmatic instance of the new mode of philosophy is Spengler's Decline of the West, a gloomy 1918 publication which, says Forman, fit the time: "...on the whole its fatalistic-relativistic pessimism was precisely the right tone for a defeated Germany." (p. 30) According to Forman,

The Spenglerian account of world history is based on the proposition that the principal cultures are autonomous organisms, each wholly unique apart from a common life cycle. Every cultural manifestation - art, science, or whatever - is simply and solely an expression of the soul of that particular culture and as such is neither "valid" nor even comprehensible outside that culture... (p. 31)

Moreover, Spengler maintains that over the course of the twentieth century the "Destiny Idea" would replace the "Causality Principle". Forman cites this from Spengler:

Destiny is the word for indescribable inner certainty. One makes the essence of the causal clear by means of a physical or epistemological system, by means of numbers, by means of conceptual analyses....The one requires us to dismember, the other to create, and therein lies the relation of destiny to life and causality to death. (p. 33)

It seems rather difficult to see what the difference between destiny - which utterly and irrevocably determines the future - and causality - which utterly and irrevocably determines the future - might be, except that destiny is transcendently good and causality the putrescent core of evil (and physicists the wicked bearers of death). (Perhaps it is the difference between efficient and final causality.) But we must also bear in mind that, according to Spengler, since I write this in a culture that is not Weimar Germany, I work a different philosophy and a different history from that which is German and hence cannot understand his work.

Sarcastic critique aside, it is crucial to note another of Forman's citations from Spengler: "The goal [of science] reached, the vast and ever more meaningless and threadbare fabric woven by natural science falls apart. It was, after all, nothing but the inner structure of the mind ..." (p. 37, emphasis added) According to Spengler, physicists do nothing but spin out the nature of their own subjectivity, as in Kant.

Forman notes that "...perverse and denigrating as this image is, it must still be recognized for what it is - an integral part of an analysis of Western culture, its present state, and its future prospects, which expressed and shaped the notions and inclinations of the educated middle classes in postwar Germany." (p. 36) Moreover, the book had sold between 1923 and 1926 100,000 copies even though there were only 300,000 college graduates in the whole of Germany. The book was "Almost universally read..." (p. 30)

Over time, central European physicists came to internalize these views. While physicists at first began only to recast the purpose of their discipline into one more humanistic, they slowly began to give up causal explanations and indeed violently reject causality, long before there was any motivation from quantum mechanics to do so. A paradigm case of this process was the conversion of Wilhelm Wien, rector of the University of Würzburg from 1914-1928. Just before the 1918 collapse, Wien gave a public lecture to the effect that physics was utterly independent from philosophical concerns. He invoked Helmholtz's name and referred to him as a "pure empiricist in particular opposition to the German idealist philosophy, above all to Hegel." (Forman, p. 41) He spent most of his time defending a modified version of Mach's extreme positivism.

In a 1919 article commemorating the 25th anniversary of Helmholtz's death, Helmholtz was recast as an opponent of Hegel - not idealism - and one with 'ideal' rather than 'material' goals in mind for science. Forman notes that

Wien thus implicitly concedes the series of equations made repeatedly by the antagonists of modern science - empiricism=positivism=narrow specialization= utilitarianism=materialism - attempting to make only an exception of Helmholtz, who, we are assured, were he alive today would look to "German idealism" to put us on our feet again. In this essay there is, of course, not the faintest touch of positivism; no mention of Mach at all. (p. 41)

In a 1920 address, he "...represented the 'postulate of the cognizability of nature' as 'in the final analysis not so very far from the fundamental idea of the Hegelian philosophy of identity'..." (p. 42) In 1925, he wept while speaking to the effect that philosophy ought to be a subject matter on which all German Ph.D. physics candidates ought to be examined. In this same year, he stated in a lecture that "...the goal of science is culture..." (Forman, p. 44)

In 1929, the notion that physics was of value not for technological application but for satisfying human spiritual drives and cultural expression was printed into the "...opening pages of the volume on 'General Foundations of Physics' in the new edition of the Handbuch der Physik ." (p. 45)

In 1920, Richard von Mises had begun to lead the charge against science from his position as chair of applied mathematics at the University of Berlin. He argued that the technological age which had given rise to classical physics was over and that the academy ought to enter the new realms of atomic and relativistic physics. These areas are to be preferred because they have "taken up again 'the question of the old alchemists...'" (p. 49) and begun to reenumerate number mysteries. It seems that von Mises was borrowing, not from Spengler, but from Sommerfeld who had recently written an article on "A Number Mystery in the Theory of the Zeeman Effect." Sommerfeld spoke in 1925 to the effect that he was not advocating mysticism but merely opposing the empiricist trends of the previous period. But as Forman points out, "Yet it is perfectly clear that... Sommerfeld was indeed catering to the antirational as well as the antipositivist inclinations of his audience.... And one cannot help but be struck at the close correspondence between [his image of physics] and that which Spengler sketched in the final pages of the Decline of the West ." (pp. 50-51)

The physicists seemed to be inviting a sense of crisis and doom. Hermann Weyl declared "The New Crisis in the Foundations of Mathematics" which, as Forman notes, "...was precipitated virtually out if thin air in the two or three years following Germany's defeat." (p. 60) In physics, the sense of impending disaster was invested in the "crisis of the old quantum theory" which physicists felt before 1925. But those physicists who wrote of the crisis in 1921 and 1922 didn't have any common ideas about its source or nature. This suggests that the physicists were looking about for a crisis and managed to find one, not that they had uncovered some deep philosophical problem in their researches. It had become fashionable, Forman argues, for physics to be engaged in its death throes and final crisis.

Causality, of course, is the central concept of classical physics. Hence Forman notes that "...if we find physicists repudiating causality - and taking pleasure in doing so - without making any attempt to critically analyze and revise the notion itself, then I think we must construe such repudiation as directed against the sort of cognitive enterprise in which physicists theretofore had understood themselves to be engaged." (p. 66) It is crucially worth noting, also with Forman, that physicists in the 1920's

...occasionally [use] the word "causality"... in several senses narrower than, not wider than, "determinism" - as equivalent to the laws of classical mechanics, to the conservation of energy and momentum, to visualization in space and time, to the absence of action at a distance, to action by contact, or to description by differential equations. And again, in many instances these special definitions of causality were advanced in conjunction with, and as the justification for, an assertion of the invalidity of the law of causality. (p. 69)

Note that on pp. 6-7 and 18-19 I have discussed Bohr's slippery and impoverished characterization of causality; for him, causality seems to be action by contact or visualizability. And Bohr's argument against causality rests on the very impoverishment that Forman here points out was characteristic of German physics as a whole. Since I have not found any other source for this particular idiosyncracy of Bohr's thinking, it is plausible that the intellectual milieu of mid-20's physics was the influence under which he so mistreated the notion of causality. Wherever Bohr got the idea, Weimar German physicists did not get it from the physics. We need to notice that causality was being gutted as a concept long before the 1925 discovery of quantum mechanics. It is likely that this predisposition against causality is the very reason why quantum mechanics was interpreted as necessarily acausal and why Bohr's interpretation got such good mileage.

IV. What Does It All Mean?

Bohr developed a (broadly!) Kantian interpretation of quantum mechanics under the influence of philosophers reasoning in a (broadly!) Kantian way and promulgated it to a scientific community which had already accepted a (broadly!) Kantian interpretation of human reasoning as such as being a spinning-out of its own nature and not a report of an independent world. This shows more, however, than that Kant had tainted Weimar German physics. The general lesson which we may be able to draw is twofold. On the one hand, science is not automatically objective. On the other hand, that which tends to make science non-objective is subjectivist philosophy.

I have shown how Bohr, and Forman has shown how Weimar physicists in general, came to insist on an acausal and anti-realist interpretation of quantum mechanics without experimental reports themselves stimulating the interpretation. In fact, the physics community was working under philosophical influences. This introduces an indeterminacy between experimental reports and their interpretation. I can best explain this with the help of Pierre Duhem's notion of theory indeterminacy. (Duhem, Aim and Structure of Physical Science , 1905) Duhem notes that in any experimental report, the precision of the data is finite. If we trace the data as a line in an x-y coordinate grid, we would actually fill up a space with a certain width when we allow for the margin of error of the data. For instance, if x and y increase in direct proportion but with a certain margin of error, there will be a space the width of the margin of error drawn at 45 from the origin of the graph. The physical scientist, in seeking to interpret this data, provides a mathematical equation which describes it. However, there are an infinite number of equations which describe lines moving within the prescribed area. Hence, the data always fails to determine which if any physical theory or interpretation is uniquely correct, though it does serve to select group of theories which fit the data. The one of these theories which is mathematically simpler is usually the best theory.

At a higher level, the data always underdetermines ontological interpretation. Persons working under different philosophical presuppositions will interpret the data in different ways. A religious mystic sees the evidence of God's existence in every action, a Kantian sees the working of the human mind, a causal realist sees the objective world acting in a certain way according to its own natural laws. Which one of these interpretations is the true one is a question for philosophy, not for science (though scientists have as much right as anyone to speculate in these areas). Hence, what scientists report that they have seen is a matter not only of the experimental results (which don't interpret themselves), but sometimes also of philosophical influence. Objectivity in science rests in part on objectivity in philosophy. A culture, such as Weimar Germany, which abandons objectivity in philosophy in favor of a subjectivist account is very likely to 'find' evidence for this very world-view in its science. If these accounts of influence are correct, then they provide evidence for an interpretation of history in which philosophical ideas are a very powerful factor in shaping events, and hence for greater caution in the philosophical training of scientists and political leaders.

But we must avoid monistic reductionism in favor of dialectical methodology. It would be a bizarre kind of monistic historical idealism to suggest that material factors are inefficacious and that philosophy is history's sole driving force. Weimar Germany was ripe for Spengler's Kantian message because it had lost the First World War, the message did not cause the loss. Other ideas which stemmed from philosophy, such as nationalism and mercantilism, may have helped cause the war and the loss, but they did so in combination with an array of material factors such as new weapons and propaganda techniques. There is a dialectical interplay between material and ideological factors. Science is free of neither.

Since philosophy is the discipline which asks and answers the basic questions about the relation between the subject and object and the possibility of objectivity, it is no surprise that the philosophic milieu (in a material context which it forms and which forms it) should profoundly influence our interpretation of phenomena as objective or subjective. The deep lesson to draw from a dialectical study of complementarity in philosophy and history is that science describes the kind of world that philosophy tells it to describe.

Find Enlightenment at