Big, Bigger, Too Big?
From Los Alamos to Fermilab and the SSC
By Lillian Hoddeson, Pais Prize Invited Talk, Atlanta Meeting, April 2, 2012
I am honored to be awarded this year's Pais Prize, and in particular for my work on big labs, which I started more than 35 years ago as I was making a switch from physics to history of physics. I was then taking graduate courses at Princeton in history of physics while teaching physics at Rutgers. To use some solid-state physics I had studied in grad school, I decided to write one of my term papers on basic research at AT&T, a group that eventually turned into Bell Laboratories. This topic pleased my Princeton mentor, Tom Kuhn, because I was able to go on and study how the quantum theory of solids impacted Bell Labs. During this period Charles Weiner and later Spencer Weart, the first and second directors of the AIP Center for History of Physics, taught me how to conduct oral history interviews for use in studying recent physics.
My plan to write a book about Bell Labs never materialized because of two unexpected, wonderful opportunities that arose after I moved to Illinois in 1977, opportunities that led to histories of Los Alamos and Fermilab. It is largely those histories that I'll draw on in this talk, about the road that led from Los Alamos to the SSC via Fermilab. As we know, this road was eventually blocked, but is still worth studying.
Los Alamos. The story of how I came to study Los Alamos had some humorous moments even in my very first visit there, in March 1977. Soon after my arrival, I was contacted by Mike Simmons and Dave Sharp, coheads of the Theoretical Physics Division. They knew I was an historian and interested in laboratories because an article of mine on Bell Labs had just appeared in Physics Today. They had come upon a safe full of interesting wartime documents which they had been ordered to send to the National Archives. At that time Los Alamos had a records center but no archives. They were reluctant to send the safe to Washington because they had heard that this would be like sending the papers into a black hole.
I was happy to try to help, but I had no clearance. I was escorted to the safe, and had to turn my back while David Campbell, Gordon Baym, and Mike Simmons examined the documents. I listened with great interest to their comments: "Wow, here's a handwritten calculation by Bethe!" "Here's Feynman's notebook!" "This letter addressed to Henry Farmer [Fermi's code-name] starts out ‘Dear Enrico!'." My favorite overheard comment was "My God, Gamov's gin bottle!" (Gamov was known to keep that cherished object safe in the safe.) All this led eventually to the Los Alamos archives and to its history project. The first outcome of that was the technical history, Critical Assembly, which I coauthored with archivist Roger Meade and two history graduate students, Paul Henrickson and Catherine Westfall. At first Los Alamos was not sure it wanted to fund the project. The director of the laboratory, Harold Agnew, is reputed to have told Simmons, "Los Alamos needs a history project about as much as it needs topless waitresses in the South Mesa Cafeteria."
This remark is difficult to understand. Project Y, the Los Alamos lab's original name, was a very well-funded, large military research facility run jointly by Robert Oppenheimer and General Leslie R. Groves. Its goal was to build uranium and plutonium bombs in time for possible use in World War II by the summer of 1945. Most people at Project Y had a security clearance and worked "behind the fence," but Oppenheimer nevertheless shaped the lab into a research facility, with theory and experimental divisions, study groups, and seminars, because he understood that fundamental research, and especially physics, was essential to solving the technical problems. That is why people like Bethe, Feynman, Rudolf Peierls, Edward Teller, John von Neumann, Luis Alvarez, Bob Bacher, Bob Wilson, Ed McMillan, Niels Bohr, Enrico Fermi, Normal Ramsey, George Kistiakowsky, and many other leading scientists were needed there.
The original program soon grew into a much larger effort in response to a physics discovery made in April 1944 by three of Emilio Segré's graduate students. Working in a secluded New Mexico canyon, they found that in reactor-made plutonium there is a small but significant amount of naturally occurring spontaneous fission, a process emitting neutrons. As the spontaneous fission rate was five times that of the cyclotron-produced samples used until then, attempting to assemble a plutonium weapon using gun assembly—which is a slow process in comparison with the speed of a nuclear explosion—was far too risky. The extra neutrons could set off the explosion too early, causing a "fizzle." An alternative method was needed to assemble a plutonium weapon. Uranium could be, and was, assembled by the gun method; for plutonium, the only conceivable possibility was implosion, a far more rapid assembly. But it was a stretch to make implosion assembly work because the physics was poorly understood.
Groves didn't want to waste the huge investment already made in plutonium production, so he simply ordered Oppenheimer to make an implosion bomb by summer 1945. This required a total reorganization and expansion of the laboratory in the summer of 1944—just one year before the deadline. Oppenheimer added, among other things, a new explosives research division under Kistiakowsky, and a new implosion research division under Bacher. Thus what started as a small, back-burner implosion program grew into a model ``big science'' program. But the program was very different from existing physics labs, and not only because of the secrecy. First, its work required scientists to collaborate with military and engineering people. Second, it had access to effectively unlimited funding. Third, it had an extremely tight military deadline. Not science as usual! The project had to move faster, and its products, the weapons, had to work reliably. The result was a conservative and redundant research strategy aimed at avoiding risk. Both bombs—and especially the implosion bomb—turned out to be overdesigned clunkers. The strategy paid off, but the special conditions that nurtured the new approach could continue only under the unique wartime pressures.
Fermilab. My work on Fermilab also started out in an unusual, somewhat humorous way. Bob Wilson had included an archives area in the blueprints for his distinctive new Hi-Rise. His archives committee contacted Joan Warnow, the archivist at the AIP Niels Bohr Library and Center for the History of Physics, who told the committee's chairman Dick Carrigan that I had just moved to Illinois and might be available. I hadn't expected to work on history of high-energy physics—and certainly not as an archivist, for which I had no training— but I was interested and agreed to help on a part-time basis.
Not long after, Bob Wilson came by and asked how I was doing. He was unimpressed when I told him what I'd learned about computerizing collections and protecting documents. When I pulled out the new fireproof boxes I had bought, he struck a match and set one on fire, causing huge flames in the middle of the history room, which fortunately died down quickly. Wilson said, "Well good. They work!" and walked out. It took me a while to realize that what Bob really wanted was a history project, with an archives and a reading room to offer culture to his staff. That suited me fine. I was delighted and relieved when in 1983, Leon Lederman, Fermilab's second director, hired Adrienne Kolb to handle the archives. She and I have worked together ever since. Later, Catherine Westfall joined us in studying Fermilab's early history, and we all eventually collaborated on Fermilab: Physics, the Frontier, and Megascience.
Many ties connect Fermilab to both Lawrence's Lab and Los Alamos. For instance, Wilson had been one of "Lawrence's boys" when he was in grad school, and in 1967 when the 200 BeV project came to Illinois, it came from Berkeley. At Los Alamos, Wilson headed the cyclotron group and later the experimental physics division. The network became most clear to me when I interviewed Priscilla Duffield, Lawrence's secretary at LBL. She later worked in a similar capacity for Oppenheimer at Los Alamos, and then for Wilson at the National Accelerator Lab, NAL, later named Fermilab. Before Duffield, Wilson's first assistant was Rose Bethe, who had helped Oppenheimer set up the Los Alamos housing office. In one interview, Wilson told me that in creating NAL he tried to recreate a science city reminiscent of Los Alamos.
Certain approaches in building Fermilab's accelerators and other apparatus resemble those used at Los Alamos, in that they often mixed engineering and scientific approaches—but unlike Los Alamos, it had a lot to do with the fact that funds, especially in the 1970s, were limited at Fermilab. But for Wilson, frugality was not just a response to limited funding but a matter of aesthetics. He liked to design minimally, taking measured risks and generally working with the least possible amount of money. He famously wrote: "Something that works right away is over-designed and consequently will have taken too long to build and will have cost too much." Subsequently, many experimentalists at Fermilab suffered because of Wilson's underdesigned Main Ring and inadequate experimental areas. But at least they had an accelerator to work with, with which they eventually discovered the bottom quark.
One of the best examples of the mixing of experimental and scientific approaches at Fermilab was in developing the pioneering superconducting accelerator magnets for Wilson's Energy Doubler, designed to double the energy of the Main Ring. It was a good idea, but the early magnets did not work well. Alvin Tollestrup succeeded in making working Doubler magnets using a brute force approach remarkably similar to that used at Los Alamos in its implosion development, building over a hundred prototypes and changing just one attribute from magnet to magnet. The Doubler was completed under Lederman, and became the basis of the Tevatron. With it Fermilab moved on to much bigger experiments, entering the regime we called "megascience," characterized by long-lasting "strings" of experiments and their follow-ups. In a limited funding context, these strings led to a general reduction in the number of problems being studied, a trend that continued for some time. The research yielded the 1995 co-discovery of the top quark, at the CDF and DZero detectors.
The SSC. I want to preface my story about the SSC with a quote from George Eliot's Middlemarch: "In all failures, the beginning is certainly half the whole." In 1983, with encouragement from George Keyworth, President Reagan's Science Advisor, American high-energy physicists were inspired to "think big" and build the ambitious 40 TeV collider as an American project rather than as an international project as originally conceived. They were led to believe there would be "new money" for it beyond the base program. Such thinking resulted in the 1983 endorsement of the SSC by Stan Wojcicki's Wood's Hole HEPAP subpanel.
The two initial phases of the SSC, between 1983 and 1988—a feasibility workshop called the Reference Designs Study followed by a design workshop called the Central Design Group, or CDG—were directed by Maury Tigner, then widely considered to be the strongest in the new generation of accelerator builders. As these phases took place in Berkeley and as the SSC was then expected, at least by physicists at Fermilab, to end up at Fermilab, it appeared to some that history was repeating itself, especially as the URA, the consortium of research universities managing Fermilab, was also managing CDG. Frank Cole, head of Fermilab's library committee, had been involved in Berkeley's 200 BeV design study in the early 1960s and remembered the traumatic moment when Berkeley physicists learned that the machine would be in the Midwest, instead of California. Sensitive to the formal analogy between the SSC and Fermilab stories, Cole suggested in 1985 that Adrienne start collecting documents and prepare an evolving chronology about the evolving SSC for an eventual history. I subsequently joined the history part of her effort. It led to a proposal to DOE, also coauthored with Peter Galison, to fund an SSC history project. But while Alvin Trivelpiece, the Director of the Office of Energy Research at DOE had informally endorsed our idea of a history project, our proposal received no response. We didn't know yet that Trivelpiece, about whom I'll say more in a moment, was just then in the process of leaving DOE. Adrienne and I continued working, and in 1994, not long after the cancellation of the SSC, Michael Riordan asked to join with us. After succeeding in getting a substantial NSF grant in 1995 for a four-year (eventually five-year) project to write a history of the SSC, we interviewed many of the people involved in the history, collected huge numbers of documents (saving the entire archival collection of the discontinued SSC), and began writing draft chapters. But it was a much bigger story than we had realized, and finishing the research required a later NSF grant in 2008.
To get back to the CDG: like his Cornell mentor Robert Wilson, believed in building imaginative and frugallydesigned accelerators. A crucial feature of CDG's design was the boldly small 4 cm aperture of its magnets, designed to achieve 6.5 Tesla. With an accelerator ring 52 miles in circumference, the CDG-designed machine was to achieve 20 TeV in each proton beam at the cost of roughly $3 billion. CDG's frugal Conceptual Design passed its intensive Temple review on April 1, 1986, even though some worried that a 4 cm aperture might not offer adequate field quality.
By this time, however, a number of clouds hung over the SSC's horizon. First, during 1985-6 the SSC's Washington base of supporters dwindled. Keyworth, Donald Hodel, and Jim Leiss left their posts and were replaced by people far less friendly to the SSC. Wilmot Hess, in particular, who succeeded Leiss, did not work well with Tigner. Second, the funding climate worsened and hope for new money faded. Many worried that the "Gramm-Rudman- Hollings ax" might fall on their work; furthermore, Presidential approval, which was required for the SSC given its multibillion-dollar construction budget, was not yet assured. President Reagan approved the SSC only in January 1987, citing the maxim of Kenny Stabler, the Oakland Raiders' quarterback: ‘Throw deep!' Trivelpiece, who received this idiosyncratic endorsement, was by then the SSC's only remaining highlevel advocate in Washington.
A third big cloud was technical, involving shorts and quenches plaguing the development of the SSC's superconducting magnets. After the retirement of Victor Karpenko, a former Livermore engineer who was one of the directors of the CDG magnet program, John Peoples of Fermilab was called in to help coordinate magnet development, then carried out by an awkward collaboration of groups at Brookhaven, LBL, and Fermilab. Peoples recalled, "I was a little bit like the plumber who'd been called in to fix the leaks and the toilets that are overflowing during a dinner party, but I wasn't exactly invited to dinner." Peoples also noticed what he called a "philosophical problem," a misalignment between the research practices of physicists and the military engineering types who responded to orders, unlike physicists who had to be seduced to work in collaborations. At Los Alamos, social misalignments of this kind were easily overcome because everyone shared a common urgent national goal, but at the SSC that was not the case. At the same time, Tigner was constrained in his leadership because he lacked control of the purse strings in paying the different laboratories for their work, which made it hard to efficiently "harness their abilities and enthusiasms," as he once told me.
The clearest sign that the SSC would not follow the road from Los Alamos via Fermilab came early in August 1988, when the DOE issued its official Request for Proposals (RFP) to manage and operate the SSC. More than a year earlier, CDG and URA had sensed, but not understood, a strangeness in the DOE's attitude toward the SSC. CDG had submitted an unsolicited management proposal that received no response. Written by Ned Goldwasser, this earlier proposal argued for solesourcing the SSC's management to URA. Meanwhile, in April 1987, Trivelpiece left the DOE. In retrospect, we can view this response to the unsolicited proposal as an indication that the SSC would not be treated in the same way as prior high-energy physics projects had been treated. But at the time this sign was not correctly interpreted by the physicists. And there was no Leslie Groves to explain it to an Oppenheimerlike leader. As Ed Knapp, the new head of URA, saw it, DOE did not seem to trust physicists to manage a project as big as the SSC. Jack Marburger later commented that any proposal with a multibillion-dollar budget was crossing an invisible line at the DOE beyond which a more elaborate funding and management process was required. If so, the physicists' hope to manage the SSC was doomed from the start.
Knapp read the DOE's RFP as a message that responses were expected to take the form of a DoD-style proposal. To write it, he hired Douglas Pewitt, then working for the Washington-based firm of SAIC. In the proposal, Pewitt included a detailed project management plan calling for teaming between physicists and industrial firms. URA then selected EG&G, Inc. and Sverdrup Corp. CDG was excluded from the proposal writing, to avoid what SLAC director Pief Panofsky saw as a potentially disqualifying conflict in which URA might appear to be taking advantage of its close relationship. This exclusion caused mistrust and resentment, especially as many others in the high-energy community were consulted. Among the off-limit topics was selection of the SSC Director. The CDG physicists considered Tigner the perfect leader, but others worried that he might project the wrong image in Washington. When on August 28, 1988, a selection committee chose Roy Schwitters as the director to be named in the URA proposal, with Tigner listed as Schwitters' Deputy Director, it caused enormous resentment at CDG. And when Waxahachie, Texas was announced as the SSC's home, on November 10, 1988, few members of the CDG were asked to go to Texas, and even fewer elected to go there. Tigner ultimately withdrew from the project when it became clear over the next few months that he and Schwitters had too many differences to be able to work together. In losing Tigner and the CDG, the SSC, and physics more generally, suffered a traumatic loss of continuity and institutional memory.
Meanwhile, the URA's response to the RFP had become a six-inch-thick, three volume document full of technical and practical details. The character of this proposal was fundamentally different from URA's elegant unsolicited proposal of 1987. Knapp later said of URA's response to the RFP, submitted on November 4, 1988, that "bureaucratically it was gorgeous." As DOE received no other proposals, a contract with URA was drafted. With its multibillion budget the SSC takes our story of big labs back to a pattern resembling Los Alamos, but without unlimited funding and without a compelling military mission.
By this time, shrewd observers whose careers were not tied to the SSC could read the future of the SSC in the tea leaves. In late October 1988, Fermilab magnet guru Dick Lundy offered three predictions to his friend Drasko Jovanovic, who was keeping an event logbook: first, that the site for the SSC would be in Texas; second, that URA would be the SSC's M&O contractor; and third, that "the project would fold in equal to or less than five years." Lundy recognized that the SSC, because of its high cost, had moved into a funding category that demanded a project management framework more like Los Alamos but without a wartime context to justify such a framework.
In any case, the SSC experienced continuing culture clashes, not only between physicists and military engineers, but also with the increasingly bureaucratic DOE. The cost of the SSC continued to grow, due in part to more bureaucracy and to design changes that added more conservative and expensive features, like a 5 cm. magnet aperture. But unlike Los Alamos, where cost increases aimed at reducing risk were no issue, Congress saw the SSC's cost increases as the result of poor management. The project was subjected to increasingly uncomfortable public and Congressional scrutiny, while the DOE's management procedures led to the alienation and withdrawal of many of the SSC's most creative scientists. Exacerbated by criticism from scientists in other fields, who feared that the SSC would cut into their fields' funding, and who pressed for its cancellation, and by being in a post-Cold War climate in which physicists had lost much of their earlier cultural prestige, the project failed to gain international support. All of these factors made the SSC crucially different from its predecessors and sealed its doom, closing the road from Los Alamos to the SSC via Fermilab. For American particle physics the death of the SSC was a tragedy that meant years of lost time, money, effort, and emotion. There were few gains for high-energy physics. One possible one is that after the SSC died, high-energy physics research seems to have become a bit more diversified, not nearly as focused as it had been on the single problem of the Higgs particle, which Lederman had originally called the "goddamn particle," because of its elusiveness, before his publishers changed it to the more charismatic "God particle." It may be that Lederman was right the first time.
The articles in this issue represent the views of their authors and are not necessarily those of the Forum or APS.