by Peter Byrne
All the physicists come to the Institute every day. Every one of them has his own—if not very big—office. At the Institute there are also experimental laboratories…. The lunch room serves as Common Room, where everybody comes together twice a day for tea or coffee (one has to buy sandwiches in a nearby shop and bring them back). It is characteristic that besides paper napkins there are sheets of paper on the tables upon which one can write formulas.
During my stay, i.e. about a month, everyone gathered together twice for an evening’s entertainment, where one showed movies and where the participants themselves entertained (accompanied by general laughter, the theoreticians had to perform very simple experiments), and a modest evening meal was had. My work at the Institute consisted of attending other scientist’s lectures and speeches and above all conversations with Niels Bohr about the fundamental problems of quantum mechanics.
Fock admired elements of Bohr’s philosophy and was thrilled to talk with him about the foundational questions:
I have become aware that even if he is more than seventy he is spiritually young, that he can get excited and talk passionately, but that he always speaks honestly. He doesn’t try to impress you by his authority, but he is convinced that he is right and he considers a patient exposition of his point of view as a weapon in a discussion…. Above all, there was the difficulty that Bohr became so engaged in the formulation of his thoughts that it was difficult for me to enter into the conversation.
Fock’s professed admiration was counter to the official stance of Soviet physics, which considered Bohr’s artificial separation of the quantum and classical worlds to be positivism.3 Marxist-Leninists disparaged positivism because it denied the existence of a knowable, objective reality. But according to Fock, “Bohr declared from the beginning that he is not a positivist and that he simply endeavors to consider nature as it is.”4 This is an interesting statement considering Bohr’s dictum, as later reported by Petersen, “The task of physics is not to find out how nature is, but to find out what we can say about nature.”5
Fock continued:
I pointed out that many of his formulations suggested a positivistic interpretation, which obviously he did not at all wish they should. I stressed the necessity of giving all quantum mechanical concepts ‘a rational foundation’ as reasonable abstractions on the basis of his own interpretation of the experiment. He answered that in no way did he reject their lawfulness. Our points of view gradually became closer; in particular, it became clear that Bohr fully recognizes the objectivity of atoms and their properties. He realizes that one has only to neglect determinism in the sense of Laplace, but not causality in general, that the expression ‘uncontrollable interaction’ is inadequate and that all physical processes are controllable. Perhaps it should be said that the similarity of our points of view only became clear through our conversations, but that it had existed before and independent of the conversations.6
Clearly Fock considered causality to be consistent with Bohr’s interpretation;7 which might have given Everett hope that the great man would extend his generalization of causality to include the uninterrupted flow of the Schrödinger equation. In any event, he went out of his way to meet with Bohr and try to hash it out.
Dining with Bohr
Everett, Nancy, and baby Liz arrived in Denmark on March 17, 1959. They had planned a six-week European vacation, spending two weeks in Copenhagen, so that Everett could confer with Bohr. After that meeting, they were scheduled to meet up with Katharine, who was to accompany them on a tour of France. Meanwhile, they ensconced themselves at the luxurious Hotel D’Angleterre.
Misner was also in Copenhagen, spending spring and summer at the Institute, where he had already met his Danish fiancée, Susanne Kemp. Her father, a prominent lawyer, was a financial supporter of the Institute, and Bohr’s long time chum.8
Misner, Susanne, and the Everetts attended a small dinner at Bohr’s palatial home. The mansion had been built by the founder of Carlsberg brewery company, which donated its use to Bohr, who was considered to be a national treasure. The downside was that the house adjoined the brewery and stank of fermenting hops. After a salmon first course and the main dish of pork roast, Everett pulled out a cigarette. Susanne, horrified, quickly informed him that in Denmark polite people do not smoke at the table. He refrained, but in the years to come, whenever he lit a cigarette in her presence, he exclaimed, “Oh, Hugh, but we do not do that here!”
Nancy, Everett, Liz in Copenhagen 1959.
Visiting Charles and Susanne Misner in Copenhagen 1959.
Susanne recalls, “Hugh nearly got kicked out of the hotel for putting cocktails and food on the outside window sill. He was sloppy and had a cigarette all the time. He ate very rich food and was fat and never had any exercise.”
Misner adds, “And he drank a lot, too.”
Everett did not invest much time hobnobbing with the other visiting physicists. Because he hated public speaking, he did not make a formal presentation on his theory, but he pitched it in private to Bohr and four other physicists, including Misner and Rosenfeld during the course of a couple of afternoons.
Bohr and Everett listened to each other, but Bohr was very difficult to understand. Misner recalls:
He would look down and think deeply and begin a sentence and halfway through the sentence his pipe would go out and he would go to the blackboard where there was a little gadget you would expect to find in a tobacco store and he would relight his pipe and it would last for three or four sentences, and then he’d do the whole thing over again. He was hard to hear. You had to lean close.
And that was it. There was no great debate, no wielding of equations, no logical traps or conundrums about relativity and uncertainty to negotiate (as there had been between Bohr and Einstein). There was simply a polite hearing and a lot of mumbling. Misner says, “Bohr’s view of quantum mechanics was essentially totally accepted throughout the world by thousands of physicists doing it every day. And to expect that on the basis of a one hour talk by a kid he was going to totally change his viewpoint would be unrealistic.”
Nor was Everett open to abandoning his theory on the basis of Bohr’s opposition. Disgruntled, he spent most of his time sight-seeing, or drinking beer in the bar of the Hotel Østerport. And in that bar, he conceived another bright idea. It had nothing to do with quantum mechanics, but it revolutionized operations research and affected the course of the Cold War.
Looking for the magic
In the late 1700s, an Italian mathematician, Joseph Louis Lagrange invented the Lagrange Multiplier method. The Lagrange multipliers are variables signified by the Greek letter λ (Lambda). The multipliers are used to optimize our ability to predict the consequences of change. For example: a farmer wants to fence a circular pasture. He knows the number of square feet in the area of the circle. He knows that 100 feet of fence will enclose the circle. Using Lagrange’s multiplier method he can determine how much land can be enclosed by adding one foot to the fence.
Military operations researchers deal with much more complex problems than fencing in a circle—problems that have thousands of continuously changing variables representing resources and constraints that feedback on each other, such as weather patterns affecting maximum engine speeds, or the explosive yield of atomic weapons versus the hardness of enemy missile silos. Even a small change in the value of a single variable alters possible outcomes, e.g. adding one megaton of explosive yield to a weapon may be the puff that breaks the missile silo’s back. When Everett and Pugh constructed their fallout model, they used Lagrange multipliers to calculate percentages of casualties as a function of the number and yield of nuclear bombs dropped under different wind conditions.
The computations involved in designing the fallout model were relatively easy compared to the computational power demanded by the task force working on Report 50. Everett had been thinking about how to solve this formidable problem ever since working on the fallout project, but with the necessity of constructing Report 50, WS
EG researchers were desperate to find mathematical shortcuts through vast thickets of interlocking variables. With the acquisition of the IBM 650, WSEG could run huge numbers of multiplier-type calculations, but without an optimizing algorithm the computer would be trapped in feedback loops and overwhelmed by the exponential numbers of possible relationships between the variables.
The power of beer
This brings us back to Everett’s vacation in Copenhagen: One afternoon, he wrote a letter on Hotel Østerport stationery to his WSEG co-worker, Bob Galiano. The historic letter began, “While drinking beer yesterday several ideas came about our maximization problem.” He went on to describe and christen the Generalized Lagrange Multiplier Method. And he called his λs “magic multipliers.”
The magic multiplier method was a computational shortcut. It used a simple but powerful algorithm to break up extremely complex problems into sets of smaller problems. As the values of λ are changed by single units, the computation gradually converges on a optimal solution (if there is a solution). When programmed into Everett’s favorite language, FORTRAN, the multipliers find solutions that are guaranteed to be true—balancing resources and constraints.
Operations researcher Gary Lucas, who worked closely with Everett at Lambda Corporation, explains that before the Generalized Lagrange Multiplier Method was invented,
Classical Lagrange multiplier theory was great for theoretical development in classical physics, but it was not very useful for large real world problems. There was not enough paper. Everett’s multipliers moved optimization into the computer world. It is entirely a product of that world. It would never have been developed without the computer. You could therefore say that Hugh made the transition from the classical optimization world of Newton to the modern number crunching world of the modern computer. This is quite a leap!9
Although the method itself left operations researchers gasping at its simplicity and mathematical beauty, its practice was a high art. It turned out that only Everett and those who worked with him consistently enjoyed success using the new method, which, in part, depended upon making educated guesses about what λ variables to insert into the equations. One had to have a feel for the method to make it work.
Leon Lasden at the University of Texas at Austin notes that the repercussions of Everett’s breakthrough are still being felt. His revolutionary contribution, says Lasden, was extending the use of the Lagrange multipliers from problems with continuously linked variables to problems with discreet variables that “jump” around. Before Everett, these “non-linear” variables resisted algorithmic solution.10
In August 1962 Everett wrote up his invention in WSEG Research Memorandum 25. A year later, it was published in Operations Research as “Generalized Lagrange Multiplier Method for Solving Problems of Optimum Allocation of Resources.” With a bow toward his former teachers, Kuhn and Tucker, Everett noted in his introduction, “The basic theorems upon which the techniques to be presented depend are quite simple and elementary, and it seems likely that some of them may have been employed previously. However, their generality and applicability do not seem to have been well understood at present (to operations analysts at least).”
When Everett returned to the Pentagon from Copenhagen, his new tool in hand, the success of Report 50 was assured.
BOOK 7
ASSURED DESTRUCTION
25 Everett and Report 50
A thermonuclear war is not impossible. As long as the weapons exist in plenty, such a war could explode at any time—its trigger pulled by a deliberate attack on the United States, by a massive invasion of Europe, a cold-war gambit pushed too far, a ballooning little war, a nuclear accident, or the act of a reckless officer. No one of these events seems likely to occur, but when all of the possibilities are added together and then multiplied by the number of days of cold war ahead, the chances of nuclear war in our lifetime become fearfully great.
Richard Fryklund, 1962.1
The unthinkable
In the months leading up to the presidential election of 1960, Democrats accused Eisenhower of being unprepared to wage and win a nuclear war. Pressured by public opinion, and by the fact that striking second was not a viable strategy in the age of nuclear submarines and ballistic missiles, the Joint Chiefs instructed WSEG to study offensive capabilities. In September 1959, 30 staffers, including Everett, were assigned to spend the next year researching and writing WSEG Report 50, Evaluation of Strategic Offensive Weapons. Long kept under classified lock and key, redacted portions of Report 50 are now publicly available.2 These are extraordinary source documents, revealing the origins of the doctrine of assured destruction, warts and all.
The historic report illuminates dark shadows in the history of the Cold War, particularly concerning the strategic focus upon striking first, which formed the back bone of the U.S. nuclear posture for many decades.
Locked inside The Cage, WSEG researchers were given access to top secret National Intelligence Estimates, and other closely held data on the weapons capabilities of the U.S., its allies, and its enemies. According to Pugh,
Everett’s contributions were absolutely central to the effort—for without Everett’s generalized Lagrange optimization methods the whole Report 50 project could never have been attempted. So although Everett was not the ‘project Leader,’ and was not officially responsible for the project results, his presence and his contributions were in reality the critical element that made the project possible.3
WSEG’s official history notes that Report 50, “became a basic source document, used for orienting incoming officials and initiating fundamental reappraisals of ongoing defense programs.” The command and control section became “something of a ‘best-seller’ contributing to an upsurge of interest and concern in command and control.”4 No wonder: it exposed the second strike doctrine as an iron glove lacking a fist.
Decapitation
WSEG determined that the United State’s nuclear forces were vulnerable not only to a Soviet first strike, but also to a retaliatory strike, despite the fact that the U.S.S.R. seriously lagged behind the U.S. in manufacturing nuclear arms. Squelching the idea of a missile gap, military intelligence assessed that the Soviets were not interested in launching a first strike, as they were aware of their own strategic weakness, and they knew that the U.S. was poised to blow them away if provoked. Therefore, WSEG warned against predicting Soviet capabilities by a “simple mirror-imaging of our own offensive development program.”5 Unfortunately, the implementation of the preventive-preemptive orientation6 that emerged from Report 50 compelled the Soviets to mirror the first strike capability of the U.S., making the world that much more dangerous.
WSEG shocked officials by reporting there was no possible defense against submarine-launched missiles or intercontinental ballistic missiles (ICBMs) that arrived in “clusters” surrounded by clouds of decoys thwarting defensive radars. Furthermore, high-altitude explosions of atomic weapons would fry the electrical circuits of retaliatory missiles with electromagnetic pulses.7 With a 15-minute warning window, the bulk of SAC’s bombers would be destroyed in their hangers by a surprise missile attack. Report 50 made it clear that a handful of Soviet ICBMs could decapitate the U.S. high command: “[The national political and joint military command structure] … is highly vulnerable and could not be counted upon to complete its minimum essential retaliatory functions if attacked.”8
WSEG urged national security officials to construct a command and control system capable of surviving a sudden attack and mounting at least a limited nuclear response. Certainly, the public, Congress, and most government officials, had no idea that the announced U.S. strategy of massive retaliation was built on operational quick sand: The chain of command and most of the population would be dead or dying before a second strike could be launched.
Using the Everett-Pugh fallout study, the report showed that after an attack by either superpower on the other, the majority of the attacked population that survived the initial blasts wo
uld be sterilized and gradually succumb to leukemia. Livestock would die quickly and survivors would be forced to rely on eating grains, potatoes and vegetables. Unfortunately, the produce would be seething with radioactive Strontium 90, which seeps into human bone marrow and causes cancer.9
Global politics
Henry Kissinger, one of the first advocates of fighting “limited” nuclear wars, was consulting at WSEG during the writing of Report 50, which echoes his Metternichian brand of real politick by treating nations as expendable pawns in the game of global dominance. Aware of this approach, Europeans were terrified of becoming collateral damage in the superpower struggle. Only Great Britain, Italy, and Turkey allowed the U.S. to deploy air bases and nuclear weapons in their territories.
With military projection of nuclear force restricted by lack of compliant allies, WSEG was afraid that, “the United States would find itself isolated in a sea of Communist continents,” as the Sino Soviet bloc moved without opposition into developing markets.10 National determination for colonized peoples was not synonymous with America’s national interest: “Had it not been for the tide of colonial emancipation, most of Africa as well as India, Burma, Ceylon, Indonesia, Vietnam, Laos and Cambodia would today be parts of allied territory.”11 Underdeveloped countries absorbed by their own economic struggles had a “strong incentive to avoid involvement in the cold war.”12
