by George Dyson
Von Neumann did his best. “I propose to store everything that has to be remembered by the machine, in these memory organs,” he explained to Weaver in an eleven-page letter in early November 1945. “This includes the … numerical information which defines the problem, including … intermediate results, produced by the machine while it works…[and] the coded, logical instructions which define the problem and control the functioning of the machine.” He described how “a very simple code of orders is adequate to handle everything” and “can be used to route the central control through subroutines, which may be organized into hierarchies of any desired structure.” This intermingling of data and instructions, he noted, “permits modification of orders in dependence upon numerical results of calculations which are carried out in the course of the process.” Finally, he explained how coded instructions, stored in internal memory, “endow the machine with ‘virtual organs,’ i.e. they make it behave as if it possessed certain organs which do actually not exist in the physical sense.”62
Weaver lent his personal influence and support, but hesitated to commit the Rockefeller Foundation to financing a venture in direct partnership with both the military and RCA. He decided to wait and see how “this rather novel combination of Institute-University-industrial laboratory-Army-Navy” played out before adding Rockefeller to the mix. During Aydelotte’s mission to Palestine, a letter from Weaver asking for an update on the computer project was intercepted by Marston Morse, to whom Aydelotte had delegated responsibilities while away. “I want to say to you confidentially that it would be a great service to the Institute if you would continue your insistence on a budget for the new project with details for the future,” answered Morse. “A few underestimates and the whole character of the Institute might be changed to follow through. The bigger it gets the more ambiguous it is.”63 Morse’s fear was not that the computer project would be a failure but that it would be too much of a success.
Others, asked by Weaver to review the von Neumann proposal, were less kind. “Von Neumann shows some tendency to regard the problem as one which begins in the scientific stratosphere and works its way down, instead of one which begins on the ground and works up,” answered Samuel Caldwell, of the Harvard Computation Laboratory. “The relay computer contains ‘5000 to 15000 relays each.’ So what? Does von Neumann think the electronic machine will not contain thousands of something?”64
Von Neumann found an unwavering ally in Lewis Strauss, the IAS trustee, merchant prince, and rear admiral who wielded power over the Office of Naval Research. Strauss saw the merit in von Neumann’s no-strings-attached approach. After what Los Alamos had accomplished with desk calculators, what might be next? “If we devote in this manner several years to experimentation with such a machine, without a need for immediate applications, we shall be much better off at the end of that period in every respect, including the applications,” argued von Neumann. “The importance of accelerating approximating and computing mathematics by factors like 10,000 or more, lies not only in that one might thereby do in 10,000 times less time problems which one is now doing, or say 100 times more of them in 100 times less time—but rather in that one will be able to handle problems which are considered completely unassailable at present.”65
Strauss took the lure. “The projected device, or rather the species of devices of which it is to be the first representative, is so radically new that many of its uses will become clear only after it has been put into operation,” von Neumann assured him. “These uses which are not, or not easily, predictable now, are likely to be the most important ones. Indeed they are by definition those which we do not recognize at present because they are farthest removed from … our present sphere.”66
Strauss, who was preparing to leave the navy, promised not only to make sure the computer project was funded before his departure, but, as reported by Marston Morse on Christmas Eve 1945, “to obtain a Quonset hut without cost.” The funding was delivered, as promised, but not the Quonset hut. The paperwork was minimal, and a one-page budget was sufficient to secure the needed funds. When the navy raised questions over title to the computer and associated patent rights, the contract was shifted to the army instead. “Professor von Neumann and I believe,” Goldstine wrote in 1951, “that the Institute has an almost unique contract with the Ordnance Department in that the Government has, in fact, given us a grant to build a machine for ourselves.”67
“We would have, once a year, kind of a pass-the-hat session in which we would sit up in the board room in the Institute with these representatives of all these government agencies,” remembers James Pomerene, “and they would say, ‘Well, I can put in $10,000’ and another guy would say ‘I can put in $20,000.’ And one would say ‘How about you, Joe? You’re good for $30,000, aren’t you?’ We would have our $200,000 put together, and it all worked fine.”68
Half the first contingent of engineers, as well as the name MANIAC, were imported from the Moore School. “Originally we called ENIAC the ‘MANIAC’ when it didn’t work right,” J. Presper Eckert remembers. “And later they borrowed that name.”69 The IAS project combined the practical experience derived from the ENIAC with the theoretical possibilities of Turing’s Universal Machine. There was regular contact between the IAS group and their British counterparts, although the British were constrained by the Official Secrets Act, which prevented them from confirming the existence of the code-breaking computers that had been built during the war.
“Von Neumann was well aware of the fundamental importance of Turing’s paper of 1936 ‘On computable numbers …’ which describes in principle the ‘Universal Computer’ of which every modern computer (perhaps not ENIAC as first completed but certainly all later ones) is a realization,” Stanley Frankel explains. “Von Neumann introduced me to that paper and at his urging I studied it with care.… He firmly emphasized to me, and to others I am sure, that the fundamental conception is owing to Turing.”70 Von Neumann knew that the real challenge would be not building the computer, but asking the right questions, in language intelligible to the machine. For this, if not for its machine shops and laboratories, the Institute for Advanced Study, Oppenheimer’s “intellectual hotel,” was ideal.
“Johnny had by then a very definite idea of how and why he wanted this machine to function with the emphasis on the why,” remembers Klári. “He wanted to build a fast, electronic, completely automatic all purpose computing machine which could answer as many questions as there were people who could think of asking them.”71
SIX
Fuld 219
We have been trying to see how far it is possible to eliminate intuition, and leave only ingenuity. We do not mind how much ingenuity is required, and therefore assume it to be available in unlimited supply.
—Alan Turing, 1939
“THE PROSPECT of a visit from an architect usually costs Professor Veblen a day’s work and a night’s sleep,” noted Abraham Flexner, when the construction of a headquarters for the Institute for Advanced Study was first announced. Veblen and Flexner had been at odds, from the beginning, over the question of building buildings and buying land. “The way to reform higher education in the United States is to pay generous salaries and then use any sort of makeshift in the way of buildings,” Flexner had argued when the Institute first opened for business—with Veblen, Einstein, Alexander, von Neumann, and little else.1
The Institute operated out of an assortment of temporary facilities for its first nine years. “Everybody was working somewhere else,” Klári von Neumann observed upon her arrival in Princeton in 1938. “Flexner had his office in one of the buildings along Nassau Street; the mathematicians had rooms in Fine Hall, which was the University’s mathematical building; the economists had some kind of an office in the basement of the Princeton Inn; and the few archeologists who were members, essentially worked in their own homes when in Princeton and then went out ‘on location’ to dig.”2
When Flexner conceded to the construction of Fuld Hall, he registere
d one last complaint, to Veblen, that “I would far rather rent additional floor space in 20 Nassau Street and get our minds so full of the purposes for which we exist that we will all become relatively indifferent to buildings and grounds.” He then warned Aydelotte that “I am still a little uneasy in mind about Veblen, for I think he has a hankering for rooms that are unduly large.”3
Fuld Hall, named after Louis Bamberger’s sister, Carrie, and her husband Felix Fuld, was constructed in 1939. The architect was Jens Frederick Larson, who had made a name for himself designing additions to university campuses—including Swarthmore and Dartmouth—and was drawn to the challenge of designing a new institution from the ground up. Born in Boston in 1891, Larson had enlisted in the First Canadian Overseas Contingent in 1915, heading to France as a member of the infantry and working his way up to lieutenant in the artillery until he signed up for flight training, transfixed by the air battles overhead. In 1917 he became a founding member of the 84 Squadron of the Royal Flying Corps, piloting a Royal Aircraft Factory SE-5A, an experimental design that not only outperformed the better known Sopwith Camel but was easier to fly.
According to British records, Larson, known as “the Swede,” scored at least eight aerial victories between November 1917 and April 1918. According to Canadian records, the score was nine. On April 3, 1918, he downed two opponents on a single day, having met “two formations of Pfalz and V-Strutters in the clouds at 7,000 feet.”4 He led his four companions up into the clouds and then dove back down upon the Germans, who were unable to escape. After scoring one more kill, on April 6, he retired from combat and became a flight instructor in England, before returning to the United States.
The earliest floor plan of Fuld Hall, with a central common room and offices in the wings (accompanied by notes suggesting that one wing be allocated to women) was sketched in pencil on the back of a menu from the City Mid-Day Club, 25 Broad Street, New York City, for Thursday, October 21, 1937. (The lunch choices offered to the Committee on Buildings and Grounds included Blue Point oysters on the half shell for forty cents, or Cape Cods for forty-five cents.) The result, two years later, was an imposing redbrick Georgian edifice with white trim and a copper roof, its lateral symmetry culminating in a clock tower that dominated the otherwise nondescript landscape of Olden Farm. According to Robert Oppenheimer, two young boys were once overheard in conversation on the private road, now named Einstein Drive, in front of Fuld Hall:
“What’s that? Is it a church?”
“It’s the Institute.”
“What’s the Institute?”
“It’s a place to eat.”
In addition to the dining room on the top floor, Fuld Hall housed the Institute’s administrative and faculty offices, and, in the center of the ground floor, a common room with a large fireplace surrounded by leather-bound armchairs and presided over by a grandfather clock from the Bambergers’ South Orange estate. A chessboard (and later a Go board, favored by Oppenheimer’s young particle physicists) sat near the windows overlooking the Institute Woods. Fresh newspapers, including the Air Edition of the Times of London, were skewered every morning on a polished wooden rack. Afternoon tea—a ritual introduced at Fine Hall by Oswald Veblen, who, according to Herman Goldstine, “tried awfully hard to be an Englishman”—was served on real china daily at exactly three o’clock. According to Oppenheimer, “tea is where we explain to each other what we do not understand.”
Surrounded by woodlands, open fields, and private drives, Fuld Hall resembled a private sanatorium or a large European country estate. The facilities were maintained by a dedicated staff of maids, groundskeepers, and janitors, including the extended Rockafellow family, who occupied one of the former farmworkers’ houses at the end of Olden Lane. Many of the staff remained with the Institute for life. On February 16, 1946, Mrs. Alice Rockafellow, described as “reliable and dependable in every emergency of which there are many in the cafeteria,” was given a raise from seventy to eighty dollars per month. “Her salary is less than other maids as the Rockafellows get very low rent,” the faculty minutes report.5 In one of the far corners of the Woods, beyond a stand of primeval beeches, a family of subsistence farmers who had been there since the land was acquired by the Institute were allowed to remain. During the war, the Institute’s fields were planted in alfalfa, rotated with corn and other crops. The Woods were declared a game reserve in 1945, although a limited bow hunting season for deer continued, and hunting platforms can still be seen in the trees if you look up.
Lunch and dinner were served in the fourth-floor dining room at below-market cost. The menu for October 14, 1946, included “Creamed Halibut with Eggs on Potatoes,” for twenty-five cents, or “Fresh Boiled Salmon, Parsley Sauce, Potatoes,” for fifty cents. Coffee was five cents. A note in the kitchen reminded staff of the “Einstein diet: No fat; no vegetable of cabbage family or bean; nothing ice cold.” Einstein preferred his eggs boiled for four minutes, and a baked apple for dessert.
The cafeteria was managed by Alice Rockafellow. The menus (for the Einstein diet, too) were produced, on a manual typewriter, by one of Larson’s fellow World War I–era aviators, Bernetta Miller, born in Canton, Ohio, in 1884 and the fifth woman to obtain a pilot’s license in the United States. In 1912 she had demonstrated the new Blériot monoplane—being built under license by Moisant Aviation on Long Island—for the U.S. Army. “Of course, I had no illusions as to why I was sent to College Park to demonstrate the monoplane to the U.S. government officials who were exclusively devoted to the idea of the biplane,” Miller later explained. “If a mere woman could learn to fly one, so surely could a man.” She served as a volunteer on the ground in World War I, receiving the Croix de Guerre from the French government for “helping the injured in the advanced aid stations” in the Tours, Toul, and Argonne sectors. A letter of commendation from the commander of the U.S. Army’s Eighty-second Division, dated January 13, 1919, cites how “under enemy fire, she visited the front lines, carrying a supply of cigarettes and other comforts to the men.”6 She was wounded at least once.
After the war, Miller became bursar of the American Girls School in Istanbul, and then returned to the United States, serving as bursar of St. Mary’s Hall, a girl’s school in Burlington, New Jersey, until signing on as Frank Aydelotte’s personal assistant and bookkeeper in 1941. Her memoranda are strongly worded, with capital letters for emphasis, and signed with a bold, firm, hand. “No BREAD will be consumed on THURSDAYS, No FOOD STUFFS will be fried on Wednesdays, and No PIES OR CAKES will be served on MONDAYS OR FRIDAYS,” she announced in May of 1946, when food conservation due to postwar shortages was put into effect.7
“I cannot urge too strongly the urgent necessity of the Computer people taking over ENTIRELY their accounting,” she noted in a memorandum to the director on September 13, 1946. “It is swamping my office to the point where we cannot give even fair attention to Institute matters.” Miller kept meticulous records of the tea service in Fuld Hall, reporting that during the six months of the 1941–1942 academic term, 9,605 servings were consumed, at a cost of 5.2 cents in tea, sugar, cookies, and labor each. She personally led the delegation of mothers who petitioned Oppenheimer and Aydelotte to start a nursery school at the Institute. “There are now 34 children on the project of whom 15 are of nursery school age.… The matter is urgent if the parents are to have reasonable quiet at home,” she reported in September 1947, requesting permission to turn one of the visitor apartments into a school. “The confusion in a small apartment with children about is considerable as you know.”8 Crossroads Nursery School opened in 1947 and has been filled to capacity ever since. Mathematicians produce their best work at about the same time that they produce their children, and the nursery school helped keep the two apart.
Miller, who had no children of her own, “was obviously interested in women, not men,” adds geneticist Joseph Felsenstein, whose maternal grandmother was Miller’s first cousin and who was taken on family visits to Bernetta and her companion
, Betty Faville, by then retired in New Hope, Pennsylvania, when he was growing up. At the Institute, he explains, “she became one of the people who had to stand between Albert Einstein and the world.” She held Aydelotte in high regard, but disliked Oppenheimer, who fired her in 1948. “I think the man was a complete snake,” she later claimed. “But I would never say he was disloyal.”9 Einstein gave her a personal letter of recommendation when she left.
Arriving at the main entrance to Fuld Hall in 1946—when the overcrowded offices were numbered differently than they are today—the telephone switchboard was on the left. From the ground floor, in the center, a half flight of stairs led down to the common room, with high French doors looking out onto the open field that extended as far as the old Princeton-Trenton trolley line marking the edge of the Institute Woods. The main column of George Washington’s army, under General John Sullivan, had been traversing the edge of this field on the morning of January 3, 1777, when General Mercer’s party engaged the British, just on the other side of where the Institute’s Social Science Library and new dining hall now stand.
On the second and third floors of Fuld Hall, in the center, with tall ceilings, was the library, with adjacent reading and talking rooms. On the fourth and top floors were the dining room, kitchen, and board room, with a balcony and terrace overlooking the lower part of Olden Farm, toward Stony Brook. (It is from this balcony that a love-struck Meg Ryan and Tim Robbins, their paths crossed via Einstein, gaze up at the stars in the 1994 Walter Matthau film I.Q.) Offices extended into the wings on both sides of the common room and library, with the humanists (and the director) occupying the right side of the building, and the mathematicians (and Einstein) occupying the left.