(Galison’s article is worth reading in full, it’s wonderful erisology — a synthesis of two models of scientific progress: incremental empiricism (of the logical positivists) and grand paradigm shifts (of Thomas Kuhn and others).)
Experimentalists, theorists and instrument makers are all physicists but since they do different things and have different priorities they tend to develop their own separate vocabularies and value systems. They do have to interact sometimes for physics to progress — experiments must be run, technological systems must be built.
For this they need to establish common ground, a shared understanding of how the things they do together are to be done and what the words they use with each other mean. This, says Galison, is not trivial.
The logical positivists tried to define science as the accumulation of observations towards complete knowledge, and they failed in the end because they couldn’t construct a perfectly objective and unambiguous language in which to encode observations without tarring them with interpretation. That just isn’t how language works. The meaning of words and actions vary by context and is always in flux, so we can’t assume communicating across contexts is straightforward.
Galison calls the borderlands where the various brands of physicist interact “trading zones”. The concept is lifted from anthropology and means a place where cultures come together for the purposes of exchange and new intercultural practices and terms emerge. He says:
I intend the term trading zone to be taken seriously, as a social and intellectual mortar binding together the disunified traditions of experimenting, theorizing, and instrument building. Anthropologists are familiar with different cultures encountering one another through trade, even when the significance of the objects traded — and of the trade itself — may be utterly different for the two sides.
Practices and terms are assigned different meanings by each specialty and their meanings inside the trading zone are simplified local versions[3], While these “skeleton concepts” are different from the ones used internally by the participating cultures, it all works out as long as everyone understands that the trading zone is a special place with special rules.
I will argue this: science is disunified, and—against our first intuitions—it is precisely the disunification of science that underpins its strength and stability. …
In this chapter, drawing on related work in the history and philosophy of science, I will argue that even specialties within physics cannot be considered as homogeneous communities. Returning to the idea of intuition I have sketched elsewhere, I want to reflect at greater length on a description of physics that would neither be unified nor splintered into isolated fragments. I will call this multicultural history of the development of physics intercalated, because the many traditions coordinate with one another without homogenization. Different finite traditions of theorizing, experimenting, instrument making, and engineering meet-even transform one another-but for all that they do not lose their separate identities and practices. …
The criteria that divided the practitioners of theory, experiment, and instrumentation—different meetings, different preprint exchange, different journals—were the classic sociological dividers Kuhn (and many others since) productively invoked to identify distinct communities. Moreover, the experimenters and theorists often disagreed as to what entities there were, how they were classified, and how one demonstrated their existence—just the criteria Kuhn used to identify incommensurable systems of belief. … But here we can learn from the anthropologists who regularly study unlike cultures that do interact, most notably by trade. Two groups can agree on rules of exchange even if they ascribe utterly different significance to the objects being exchanged; they may even disagree on the meaning of the exchange process itself. Nonetheless, the trading partners can hammer out a local coordination despite vast global differences. … The anthropological picture is relevant here. For in focusing on local coordination, not global meaning, I think one can understand the way engineers, experimenters, and theorists interact.
Okay, but what examples does Galison write about? Here are some:
Experimentalists—and one could make a similar statement about theorists and instrumentalists—do not march in lockstep with theory. For example, the practice of experimental physics in the quantum mechanical revolution of 1926-27 was not violently dislocated despite the startling realignment of theory: spectroscopy continued unabated, as did measurements of specific heat and black-body radiation. And practitioners of these experimental arts continued, undaunted, to conduct a continuing dialogue with theorists across the great theoretical divide. Each subculture has its own rhythms of change, each has its own standards of demonstration, and each is embedded differently in the wider culture of institutions, practices, inventions, and ideas. …
Experimenters come to believe in an effect for various reasons; one is the stability of the phenomenon—you change samples, you shift the temperature-and still the effect remains. Another road to the closure of an experiment involves the increasing directness of our probing of the phenomenon. By increasing the power of a microscope, the energy of a particle beam, the disposition·of the apparatus, or the amplification of a signal, one probes further into the causal processes linking phenomena together.
The theorist’s experience is not so different. You try adding a minus sign to a term—but can’t do it because the theory then violates parity; you try adding a term with more particles in it—forbidden because the theory now is nonrenormalizable and so demands an infinite number of parameters; you try leaving a particle out of the theory—now the law has uninterpretable probabilities; you subtract a different term and all your particles vanish into the vacuum; you split a term in two—now charge isn’t conserved; and you still have to satisfy conservation laws of angular momentum, linear momentum, energy, lepton number, baryon number. Such constraints do not all issue axiomatically from a single, governing theory. Rather, they are the sum total of a myriad of interpenetrating commitments of theoretical, instrumental, and experimental practice: some, like the conservation of energy, centuries old. Others, like the demand for naturalness—that all free parameters arise in ratios on the order of unity—have their origin in recent memory. But taken together, the superposition of such constraints make some phenomena virtually impossible to posit, and others (such as the black hole) almost impossible to avoid.
Indeed, the astonishing thing about black holes is that they form (theoretically) in the face of enormous variations in the basic structure of our theory of matter. They don’t depend on the details of this or that theory of the strong, the weak, or the electromagnetic force; and to remain consistent with other observations there is practically nothing one can do with the theory of gravity that would get in the way of the formation of black holes. … This stubbornness against variation is the theoretical analogue of stability, and it is the experience of this stability that eventually brings theorists to accept such objects come what may (almost) from their experimentalist colleagues.
In our case, theorists trade experimental predictions for experimentalists’ results. Two things are noteworthy about the exchange. First, the two subcultures may altogether disagree about the implications of the information exchanged or its epistemic status. For example, as we have seen, theorists may predict the existence of an entity with profound conviction because it is inextricably tied to central tenets of their practice—for example, group symmetry, naturalness, renormalizability, covariance, or unitarity. The experimentalist may receive the prediction as something quite different, perhaps as no more than another curious hypothesis to try out on the next run of the data-analysis program. But despite these sharp differences, it is striking that there is a context within which there is a great deal of consensus. In this trading zone, phenomena are discussed by both sides. It is here that we find the classic encounters of experiment with theory: particle decays, fission, fusion, pulsars, magnetostriction, the creep effect, second sound, lasing, magnetic deflection, and so on. It is the existence of such trading zones, and the highly constrained negotiations that proceed within them, that bind the otherwise disparate subcultures together.
The example of relativistic mass is an appropriate place to start because over the last thirty years it has become the locus classicus for discussions of meaning incommensurability. For Kuhn, the advent of Einsteinian dynamics was a prototype of revolutionary change and, he argued, only at low velocities could the two concepts of mass be measured in the same way. On this view, one would expect there to be no experimental mode of comparison of Einstein’s concepr of mass and the concepts of mass his theory displaced—those of H. A. Lorentz, Max Abraham, and Henri Poincare, none of whom shared Einstein’s view of an operationally-defined space and time. … Kuhn’s claim is that prerelativistic and relativistic uses of the term mass make comparison impossible: “Only at low relative velocities may the [Newtonian and Einsteinian masses] be measured in the same way and even then they must not be conceived to be the same.”
In fact, there was a rich experimental subculture preoccupied precisely with comparing these different theories—and not at low velocities. With Max Kaufmann and Alfred Bucherer leading the way, these experimenters produced experiment after experiment using magnetic and electric fields to measure the mass of the high-velocity electron perpendicularly to its velocity. Moreover, their efforts were clearly understood by all four of the relevant theorists (Poincare, Lorentz, Abraham, and Einstein) to arbitrate among theories. Lorentz recognized the relevance of one such set to his work and immediately conceded defeat: “Unfortunately my hypothesis [explaining mass by] the flattening of electrons is in contradiction with Kaufmann’s results, and I must abandon it. I am, therefore, at the end of my Latin.” These are not the words of someone for whom the experiment was irrelevant or incomprehensible. Only slightly less despairingly, Poincare conceded that at “this moment the entire theory may well be threatened” by Kaufmann’s data. Einstein himself was more confident of his theory, and doubted the execution of Kaufmann’s work; he did not challenge the relevance in principle of the results. Quite the contrary: Einstein went to considerable pains to produce predictions for the transverse mass of the electron so that Kaufmann and Bucherer could use their experimental methods to study the theory; he constructed a detailed analysis of Kaufmann’s data; and he even designed his own modification of the electron-deflection experiments which he hoped someone would execute. …
The lesson I want to draw from this example is this: despite the “global” differences in the way “mass” classifies phenomena in the Lorentzian, Abrahamian, and Einsteinian theories, there remains a localized zone of activity in which a restricted set of actions and beliefs are deployed. In Kaufmann’s and Bucherer’s laboratories, in the arena of photographic plates, copper tubes, electric fields, and in the capacity of hot wires to emit electrons, experimentalists and theorists worked out an effective but limited coordination between beliefs and actions. What they worked out is, emphatically, not a protocol language—there is far too much theory woven into the joint experimental/theoretical action for that. Second, there is nothing universal in the establishment of jointly accepted procedures and arguments. And third, the laboratory coordination does not fully define the term mass, since beyond this localized context the theories diverge in a myriad of ways. Theorists and experimentalists are not miraculous instantaneous translators and they are not “mere” instrumentalists uninterested in interpretation. They are traders, coordinating parts of interpreted systems against parts of others.
One more example:
At first glance, the war would seem to have made no contribution whatsoever to such an abstruse and abstract subject as quantum electrodynamics. The usual story about QED runs roughly as follows: during the 1920s and 1930s physicists interested in the subject, including Victor Weisskopf, H. A. Kramers, J. Robert Oppenheimer, Niels Bohr, Julian Schwinger, and others made halting progress in understanding how the quantum theory of the electron could be combined with special relativity. They made only intermittent progress, limited essentially to first-order calculations. For reasons of war work, all those living in the United States supposedly broke off their efforts duting World War II to do their required (but “irrelevant” to pure physics) work on engineering, and then returned, triumphantly, to QED in the second half of the 1940s.
The story is false on at least two levels. First, as Silvan Schweber has pointed out, the developments in QED were catalyzed in part by the results of wartime microwave technology that made possible the precision measurements of Willis Lamb, R. C. Retherford, Henry Foley, J, M. B. Kellogg, P. Kusch et al. in Rabi’s laboratory and the work of Dicke at Princeton. These were extraordinary experiments, but the impact of the war went even deeper. Radar work reconfigured the strategy by which Schwinger approached physical problems. Schwinger himself has alluded briefly to his judgment that his radar work had a strong impact on his postwar thinking; in what follows I will expand on his later remarks, making use of his actual work in radar to complete the picture.
Let us attend to practice—not results. During the war, Schwinger worked in the theoretical section of the MIT Rad Lab; his group had the task of developing a usable, general account of microwave networks. Ordinary network theory—the theory of radio waves in resistors and capacitors—utterly failed because microwaves have a wavelength of the same size as ordinary electrical components. In ordinary components such as resistors, copper wires, or cylindrical capacitors, the microwave energy would radiate away. This meant that the full set of calculational tools available for electronic circuits became useless. With the help of his coworkers, Schwinger began with Maxwell’s equations and derived a set of rules by which engineers and physicists could malce practical network calculations.
As the war progressed and Schwinger assimilated more of the engineering culture of the Rad Lab, he began to abandon the physicists’ abstract scattering theory of electromagnetism, and to search for the microwave analogue of the electrical engineers’ more practical representations: simple “equivalent circuits” that imitated just the relevant aspects of the components. It was an old technique among electrical engineers, who were used to treating certain systems, such as loudspeakers, not by their real electrical, mechanical, or electromechanical properties, but as if the loudspeaker were a circuit of purely electrical components. In other words they (symbolically) put the complicated physics of the loudspeaker’s electromechanically generated noise into a “black box,” and replaced it in their calculations with “equivalent” electrical components. Similarly the conducting hollow pipes and cavities of microwave circuits could be replaced (symbolically) by ordinary electrical components, and so make the cavities amenable to algebraic manipulation—without entering each time into the details of complex boundary-value problems for Maxwell’s equations. As the postwar Rad Lab “Waveguide Handbook” put it, the adoption of equivalent circuits “serves the purpose of casting the results of field calculations in a conventional engineering mold from which information can be derived [sic] by standard engineering calculations.” It is just this process of appropriation—this “casting” into an “engineering mold” that intrigues me. In this detachment of field calculations from their original context, the full meaning of the terms is cut short. Nor is the meaning suddenly and of a piece brought into engineering lore: microwave frequencies did not allow any simpleminded identification of electrical properties with the well-known categories of voltages, currents, and resistances. The product of this labor was a kind of simplified jargon binding elements of field theory talk with elements of engineering equivalent-circuit talk.
In short, the war forced theoretical physicists—such as Schwinger—to spend day after day calculating things about devices and, through these material objects, linking their own prior language of field theory to the language and algebra of electrical engineering. Modifying the theory, creating equivalent circuits for microwave radiation, solving new kinds of problems was not—and this is the crucial point—a form of translation. Even Schwinger’s “glossary″ identified newly calculated theoretical elements with recently fabricated fragments of microwave circuitry; neither was part of the prior practice of either the theorists or the radio engineers. Boundaries are substantial, translation is absent, and Gestalt shifts are nowhere in sight.
Schwinger himself has alluded to the link between the two seemingly unrelated domains of waveguides and renormalization. “[T]hose years of distraction” during the war were more than that: “[t]he waveguide investigations showed the utility of organizing a theory to isolate those inner structural aspects that are not probed under the given experimental circumstances …. And it is this viewpoint that [led me] to the quantum electrodynamics concept of self-consistent subtraction or renormalization.” With an understanding of Schwinger’s work in waveguide physics, we are now in a position to unpack this connection between the calculations of radar and renormalization.
In the microwave case, it was impossible to calculate fully the field and currents in the region of the discontinuity; in the quantum electrodynamics case, it was hopeless to try to pursue the details of arbitrarily high-energy processes. To attack the microwave problem, Schwinger (wearing his engineering hat) isolated those features of the discontinuity region’s physics that were important for “the given experimental circumstances”—for example, the voltages and currents emerging far from the discontinuity. In order to isolate the interesting features, he dumped the unneeded details of the electrodynamics of the discontinuity region into the parameters of an equivalent circuit. Faced with the fundamental problem of quantum electrodynamics, Schwinger concluded in 1947 that he should proceed by analogy: one had to isolate those features of the physics of quantum electrodynamics that were important for the given experimental circumstances—for example, magnetic moments or scattering amplitudes. To separate these quantities from the dross, he dumped the unneeded details of high-energy interactions into the renormalization parameters.
One lesson that theoretical physicists learned from their engineer colleagues during the war was, therefore, simple yet deep: concentrate on what you actually measure, and design your theory so that it does not say more than you must to account for these observable quantities. The adoption of this positivist attitude toward theorizing was such a sufficiently sharp break with earlier traditions of theory, that some of Schwinger’s contemporaries never accepted it. Even Dirac, one .of the greatest of twentieth-century theorists, resisted the idea of renormalization until his death in the 1980s. But the idea rapidly took hold, altering for at least several decades the theorists’ attitude toward the limits of their description of nature.
Last quote (this “shortform” is clearly no longer short):
Despite this radical difference in the ontology—the set of what there is—a meeting ground exists around the description of the phenomenology of particle physics: How do photons recoil from electrons? How do electrons scatter from positrons? How do photons create pairs of electrons and positrons in the near presence of a proton? For these and similar questions, the experimentalists and theorists come to agreement about rules of representation, calculation, and local interpretation. In a strong sense, Bjorken and Drell Volume I is an example of an attempt to create a stable pidgin language, designed to mediate between experimentalist and theorist. Reduction of mathematical structure, suppression of exceptional cases, minimization of internal linkages between theoretical structures, removal from a more elaborate explanatory structure—these are all ways that the theorists prepare their subject for the exchange with their experimental colleagues. I take these moves toward regularization to be the formal-language analogues of phonetic, morphological, syntactical, and lexical reduction of natural languages. …
And indeed there is, not surprisingly, a corresponding “foreigner talk″ that experimentalists develop on their side. Just as theorists reduce the complexity by suppressing the “endogenous” structure linking theory to theory, so experimentalists, when addressing theorists, skip the connecting details by which experimental procedures bind to one another. These “separable” bits of procedure can come as isolable fragments of craft or engineering knowledge, as when the Alvarez group introduced indium as the binding material by which bind bubble chamber glass to the steel chassis. Between such localized wisdom and material lay computer programs such as the PANG or KICK. Their exchange not only regularized practices in the “image” tradition, the track analysis programs carried over as well into the “logic” tradition, serving in the long run to facilitate the coalescence of the two previously competing cultures.
From John Nerst’s All the World’s a Trading Zone, and All the Languages Merely Pidgins:
Peter Galison’s Trading Zone: Coordinating Action and Belief begins dramatically:
Okay, but what examples does Galison write about? Here are some:
(That last paragraph is the only way I can make sense of string theory devotees’ real beliefs.)
Back to the physicist subculture trading zone:
One more example:
Last quote (this “shortform” is clearly no longer short):