The so-called Copenhagen interpretation of quantum physics has been difficult if not impossible to pin down, and turning to the writings of its founders for clarification can prove to be disappointing. Although Bohr, Heisenberg, and Pauli presented a united front in discussions of interpretation, they also disagreed on a number of fundamental issues, including the meaning of electron spin and our ability to understand atomic structure in terms of the motions of particles. In Quantum Dialogue Mara Beller argues that we cannot hope to find any consistent over-arching philosophy among these scientists.

Beller's novel and insightful analysis about the Copenhagen interpretation comes through an application of what she calls the “dialogical approach” to the history of science. Beller argues that to understand the history of science, one must realize that all science is an open-ended and flexible process that is more like a continuing conversation than a single-minded pursuit of paradigmatic goals. According to this dialogical view, neither scientific communities nor individual scientists maintain any core set of ideas, principles, or methods that can be viewed as a Kuhnian paradigm. Rather, since scientific creativity is driven by dialogue, the various individual strands of creative exchange in a community, which themselves are associated with different social and political contexts, can lead to the adoption of disparate arguments and techniques. Scientists collaborate and react to new developments, and in response to questions or criticisms they sometimes change their minds and sometimes adopt inconsistent ideas for rhetorical purposes. Even in a single work one may find different “voices” associated with the way the author is engaged with various concerns and interests of other members of the scientific community. Though this may be especially true during the period of a major scientific breakthrough, Beller argues that it is the norm.

Of course, most historians recognize that the scientific and social contexts of a scientist's work can inform our understanding of it in various ways. Beller's approach takes this idea a step further by claiming that understanding these contexts is fundamental to the understanding of science, because the way scientists respond to these contexts forms the very essence of the scientific process. In this dialogical process, social and political factors can be predominant or “virtually non-existent,” depending on the local situation. But dialogism is not social constructivism, because it denies the existence of robust consensus among the scientists and thus contains no consensus theory of truth (322). Dialogism is essentially a description of scientific creativity in terms of the way a scientist is inspired by others’ work or personal communication, and this process neither necessarily rules out social or political factors nor necessarily incorporates them. As such, the dialogical view serves as both an historiographical approach and a description of the way science proceeds. As an approach to history, it urges us to look for the different voices arising from various dialogues with other scientists and not to try to force scientists’ views into a consistent overarching framework. As a description of scientific process, it suggests that there really aren't any consistent frameworks for us to find anyway.

Quantum Dialogue explains and endorses Beller's vision of a dialogical approach to the investigation of science by providing a number of insightful studies of the history of quantum theory and the associated philosophical debates of that period. In the first part of the book Beller argues that an understanding of the dialogical emergence of matrix mechanics and it subsequent interpretation by Heisenberg and Bohr sheds light on otherwise unexplained inconsistencies in their work. For example, she makes a convincing case in chapter three that positivistic philosophy did not drive the development of matrix mechanics. Beller argues that Heisenberg's positivistic justification of the goal of the elimination of non-observables was only an “ex post facto” defense against the perception that the mathematics of the theory was too abstract. When Schrödinger's theory began to emerge as a more promising alternative because it was easier to use and better able to be understood along classical lines, Heisenberg and the Göttingen group first extolled unintuitivity as a positivistic virtue and only later re-introduced the particle picture as the way to understand the meaning of matrix mechanics. Other intriguing parts of the book include analyses of the emergence of Heisenberg's uncertainty paper and Bohr's complementarity lecture that indicate the roles other, sometimes less-well known scientists had in these developments via their exchanges with the principal actors.

In the second part of the book Beller argues that the Copenhagen argument about the inevitability of their interpretation was merely a rhetorical strategy aimed at disguising the multiplicity and shaky philosophical foundations of their views. The main target of this analysis is Bohr, and here Beller pulls few punches. Bohr after the development of quantum mechanics is painted as a tragic figure, unable or unwilling to understand the mathematics of the new formalism, resting on his laurels, and using his father-figure status to force consensus in spite of the incomprehensibility of his ideas. It is unfortunate that Beller's own rhetoric in these sections on Bohr detracts from the insight she provides elsewhere, and especially from her analysis in chapter six on the “Como” lecture where Bohr introduced complementarity.

Beller's argument in this second part of the book is based mainly on a survey of the differences within the Copenhagen crowd as well as her previous analysis of the dialogical emergence of Bohr's and Heisenberg's interpretations. Because it relies on the interpretation of the differences between these views, Beller's argument here raises the question of the relative significance of apparent disagreement between scientists who we assume are working in the same paradigm or research program, as compared with disagreements between the larger groups or programs. Take, for instance, the Bohr-Heisenberg debate over the meaning of Heisenberg's uncertainty principle. In chapter six, Beller takes this debate as evidence that Bohr and Heisenberg disagreed about how to interpret quantum mechanics: Heisenberg provided an analysis first in terms of particles and the disturbance of the particle upon observation, and Bohr's correction was an argument for a wave interpretation. Apparently, then, they shared little in common regarding interpretation. But how are we to compare that debate with the debate between Bohr and Schrödinger over quantum jumps and stationary states? Are the differences between their views comparable in magnitude to the differences between Bohr's and Heisenberg's? Beller's claim is that they are. Thus from the perspective of Beller's dialogical framework, we need additional explanation for why Bohr and Heisenberg joined forces despite their apparent differences. But Beller's answer that it was because of the threat of the opposition seems to presuppose an idea of a common opposition, which itself suggests that they may already have believed that they had certain ideas in common. Perhaps then it is not so wrong to say that Bohr, Heisenberg and others were all working in something like a shared research program.

A potential problem with the dialogical approach as historiography is that it may make it too easy to read inconsistency into a work and to dismiss common themes throughout an individual's or group's body of work, because it sets up the expectation that one will find conflicting voices. Beller's treatment of Bohr may be a case in point. For instance, she downplays Bohr's claim that the correspondence principle continued to play an essential role as an interpretive tool after the development of quantum mechanics. This claim is associated with the importance of the use of the correspondence principle in the development of quantum mechanics, and Bohr spends much of the Como lecture (where he introduces the concept of complementarity) recounting this history. Beller notes the significant historical content of the Como lecture, but attributes Bohr's account of the historical role of the correspondence principle to the Bohr-Heisenberg dialogue: Heisenberg had failed to credit the importance of Bohr's correspondence principle in his work on matrix mechanics, and Bohr was putting Heis enberg in his place (140). She then discounts the possibility that Bohr and Heisenberg had much in common by painting their approaches as coming from the opposite directions of particle and wave conceptions of matter and radiation. Yet I would argue that there are elements of Heisenberg's reasoning that can only be understood in terms of correspondence arguments based on the wave theory of radiation. And for his part, Bohr was not entirely in favor of waves but rather treated concepts of both waves and particles as abstractions whose use in any particular situation was based on a new understanding of the correspondence principle. In fact, the historical content of Bohr's Como lecture can be understood better in terms of an argument for the inevitability of the form and interpretation of quantum mechanics, given the framework established by the correspondence principle. In such a context Bohr's argument for complementarity relies partly on the formalism of quantum mechanics, and so may not be so drastically different from Heisenberg's view of the uncertainty principle. I offer these possibilities here without supporting details simply as a suggestion that there still may be a fresh and accurate way to understand Bohr. Beller is right to point out that there has been little consensus on how to understand Bohr. But her approach, by focusing on conflict rather than continuity, seems to presume the impossibility of re-discovering within Bohr's work anything of interest to contemporary philosophy of science. The question is how deep the apparent inconsistencies in Bohr's work are; Beller's account of science in dialogue suggests that the inconsistencies go all the way down.

Beller claims that one of benefits of dialogism is that it offers a philosophy of scientific methodology and progress that overcomes difficulties that other accounts have in connecting methodology and goals (chapter 15), because the goal of strong support of a scientific claim is achieved by exposing the claim to the possibility of confirmation or refutation in the process of scientific dialogue. Unfortunately, Beller doesn't develop this connection between the historiographic claims and the philosophical implications of the dialogical approach as much as she might have. Many ideas that have informed contemporary historiography of science have been associated with a specific philosophy of scientific progress. The process by which a historian selects what needs explaining and what doesn't is based at least partly on ideas about what counts as normal (or even rational) scientific behavior. Different philosophical views of the progress of science have explicit or implicit implications for what a historian ought to attribute to an “external” influence on a scientific development, and these different views would suggest different focuses in one's approach to history. Popperian falsification claims that it is irrational to continue to believe in a theory after a negative result in a crucial experiment, and hence suggests that such a continued belief must be explained by other factors. A Kuhnian account of paradigm shifts denies the existence of crucial experiments as tools for rational adjudication between theories, and this implies that a continued belief in a “falsified” theory requires no special explanation. Lakatos’ methodology of research programs suggests that while it is not irrational to continue to hold on to the “hard core” of a research program in the face of failures, we must be honest about whether the program is progressive or degenerating; lack of such honesty is then the irrational or unscientific act that requires extra explanation. Each of these views provides some account of scientific progress to help us understand the acts of scientists. They may also be able to account for scientific inconsistency. Lakatos’ account, for one, can account for the existence of inconsistencies in models even within a research program that contains a uncontested hard core, because of the way it describes the process of the generation of models using ideas outside the core. It can also account for the existence of consistencies even in the face of an ad hoc, reactionary process of model development. The telling difference between science in dialogue and a Lakatosian research program that generates significantly varied theories is not whether theories are generated and argued in dialogue but whether the community holds any common core beliefs.

For Beller, it is ironically the appearance of consistency that requires explanation, since on her account the piecemeal adoption of inconsistent or loosely related ideas ought to be the norm. Individual words and acts apparently derive meaning more from their historical connections to the web of dialogue than from their logical relation to the phenomena the scientists are attempting to explain. Beller even compares dialogism in science to an art historian's claim that in art “all pictures owe more to other pictures than they do to nature” (315). Then what, if anything, distinguishes science from other creative activities? While Beller does not discount the role of experiment in science, she claims that the theory-ladenness of observation, and especially scientists’ consequent ability to re-interpret data in order to overcome apparent falsifications of their theories, indicates that the more important aspect of science is the dialogue between individuals that drives such reinterpretations. The challenge for dialogism will be whether it can incorporate the insights Beller has provided about scientific creativity and dialogue into an account of what makes science scientific and what distinguishes the history of science from other histories. That is, can we give an account for why this process of scientific dialogue results in theories that explain, predict, or meet some other proposed goal of scientific inquiry? Or rather, ought theories themselves be taken to be inconsistent collections of local explanations?

In short, the details of Beller's historical work are wonderful, though controversial. Beller's focus on the way various strands of scientific thought are woven together in the dialogue between scientists makes for a rich understanding of the development of quantum physics. But I am less satisfied with the more general conclusions Beller reaches with the application of the dialogical approach. We must ask ourselves what we ought to do as philosophers or historians of science when faced with difficulties in the interpretation of some figure's work. Clearly, hero-worship and assumptions of hyper-rationality can cloud our judgment, and this book offers an antidote to those tendencies by showing how varied the creative sources for and targets of a scientific argument can be. It also leaves us with some unanswered questions about the antidote's possible side-effects. These questions are well worth studying.