Grace Hopper and the Invention of the Information Age
Page 13
The Mark I manual served as a manifesto for automated computation after the war. Hopper produced an account in which the machine represented both a technical and a conceptual break from the past. This break is captured by the machine’s official name, the Automated Sequence Controlled Calculator. A fully automated machine would free itself from the limitations of the human mind. As Aiken and his team proved during the war, a fully automated machine could operate 24 hours a day, seven days a week. The automation of mathematics and information processing would subject society to unprecedented technological forces. Just as the steam engine became the technological foundation of the industrial revolution, the Automated Sequence Controlled Calculator could become the technological foundation for a new type of revolution dealing with information.
From 1945 to 1947, the popular press embraced the Aiken-Hopper version of history and propagated it.61 The successful spread of their account is best represented by the Boston Sunday Post’s cover story of 28 December 1947. Titled “Professor Aiken of Harvard Computation Laboratory Talks of Wonders / Says Era of Mechanical Calculators Lies Ahead of Us,” the article includes a picture of Richard H. Babbage, great-grandson of Charles Babbage, talking with “Professor Howard Aiken, father of the modern calculating machine” while they inspect a piece of Babbage’s Difference Engine. The article includes a condensed version of Hopper’s history that recapitulates text found in the Mark I manual almost verbatim.
The academic press recognized Hopper’s version of events just as quickly. In 1946, in the British scientific journal Nature, the British mathematician Leslie J. Comrie published a review of the Mark I manual titled “Babbage’s Dream Comes True.” Comrie concludes that Aiken’s machine “is the realization of Babbage’s project in principle,” though he is “astonished” by the omission of IBM from the title and from Aiken’s preface.62 As a leading proponent of large-scale computational machines in Britain, Comrie openly lamented the fact that the British government had failed to bring Babbage’s machines to their successful conclusions, which had “cost Britain the leading place in the art of mechanical computing.”63
Aiken and Hopper also published three articles in Electrical Engineering, the primary journal of the American Institute of Electrical Engineers. Written once again by Hopper, the articles summarize the Mark I manual.64 Their purpose was not only to educate a wider audience about computers but also to highlight the heroic achievement of Aiken and his crew deep in a basement at Harvard during the war. While some (at IBM and elsewhere) took issue with Aiken’s depiction as a heroic inventor, Hopper—who believed that his forceful personality had helped to launch the computer age—defended her former boss. Reflecting years later on her time at Harvard, Hopper said:
If somebody has imagination and thinks up new concepts and new ideas, we are very apt to put a committee around them and cut them down to size. We don’t tend to grow any giants anymore; we try to cut them back to the average to make them like everybody else. We are afraid of giants, so we are not growing any. We need a few giants.65
Hopper’s public defense of Aiken and his achievements after the war enhanced her position at Harvard and her position in the nascent computing community. From the day she was assigned to Harvard, Hopper’s instinct was to work hard and make her mathematical and writing skills indispensable to Aiken. Regardless of her military rank, Hopper was faced with the challenge of how to gain power from the margins, especially in view of Aiken’s open distaste for being assigned a female officer. By embracing Aiken’s embryonic vision and nurturing it with skill and care to maturity, Hopper endeared herself to her boss while creating a foundation for his postwar career as well as hers.
SPREADING THE WORD: HARVARD’S POSTWAR PR EFFORT
Hopper’s Manual of Operation and her articles captured the technical achievements of the Harvard Computation Laboratory. These documents, combined with Aiken’s library of photographs of the machine and its crew, were distributed to reporters, engineers, scientists, and businessmen who were interested in the field of large-scale computational machines. The documents outlined the technical aspects of the machine and placed Mark I, its inventor, and the Harvard Computation Laboratory at the dawn of the computer age.
With the conclusion of the war, Aiken encouraged dissemination of information concerning his machine. Though he had forfeited any patent rights to IBM in exchange for funding, Aiken was never one to focus on the machine’s commercial applications. Much like John von Neumann, Aiken believed that “essentially the primary object is to promote science.”66 In this light, the commander of Mark I believed the spread of computing technology to be a public good. His views were incorporated into the postwar contract with the Navy Bureau of Ordnance, which stated that no attempt would be made to restrict publication of results except in certain instances involving national security. This included information concerning the proposed Mark II and Mark III machines.67
To facilitate the flow of information, Aiken opened the doors of the Computation Laboratory to interested visitors. “It was 1946,” Hopper recalled, “and we began to do more teaching. We had a little lecture room, and we began to have more visitors.” With her academic experience, Hopper was the natural choice to lead private tours and give lectures. In the postwar environment, Hopper found herself spending a significant amount of time escorting important visitors and explaining to them how the new computing machines could be applied to a variety of fields. By the fall of 1946, Hopper was more concerned with public relations and the management than with writing code herself.68
6 THE 1947 HARVARD SYMPOSIUM ON LARGE-SCALE DIGITAL CALCULATING MACHINERY
By the beginning of 1947, Howard Aiken had taken his place among the elite of the budding computing community, and his staff, including Grace Hopper, basked in his glow. The newly minted Harvard Ph.D. who in 1940 had been informed by President James Conant that he would not be able to advance above faculty instructor was now a permanent fixture on the Harvard campus.1 Aiken utilized the organizational and financial weight of the Navy to push a reluctant university to the forefront of mechanical computation and applied mathematics.
Harvard’s heightened role within the computing community was physically represented by the construction of a modern two-story brick-and-glass building to house the new Computation Laboratory. The Laboratory was envisioned as the first of a number of new buildings that would make up Harvard’s “Science City.”2 Given the emergence of science and the newfound relationship between universities and government in the postwar era, the Harvard administration hoped to improve the school’s reputation in the sciences. In response to the new conditions, the Department of Engineering Sciences and Applied Physics was formed in 1946.
The Harvard Computation Laboratory featured on the cover of the Journal of Applied Physics, October 1946. Courtesy of Archives Center, National Museum of American History, Smithsonian Institution.
Aiken retired from active military duty on 1 January 1946, but retained his position as head of the Computation Project at the insistence of the Navy’s Bureau of Ordnance. As a condition of Research and Development Contract 8555 between Harvard University and the Navy, Harvard agreed to make Aiken an associate professor of engineering mathematics within the new department. The contract also gave Aiken authorization to reassign wartime military staff members to the Harvard University research faculty. Aiken chose Grace Hopper, Richard Bloch, and Robert Campbell as the chief deputies of his growing Crimson fiefdom.3
SHARING KNOWLEDGE
In the years 1944–1947, the Harvard crew had demonstrated that a calculating machine could produce timely, accurate mathematical solutions on a continuous basis. Farther south, Presper Eckert, John Mauchly, and their team of female operators proved the feasibility of calculation at the speed of electricity. These early successes generated a wave of new projects including BINAC and UNIVAC (Electronic Control Company), Mark II and Mark III (Harvard Computation Laboratory), SSEC (IBM), Whirlwind (MIT), EDVAC (University of Penns
ylvania), MANIAC (Institute for Advanced Study), and ERA (Engineering Research Associates). In England, the National Physical Laboratory’s ACE and the University of Manchester’s Mark I were under way, and at Cambridge University Maurice Wilkes was beginning his plans for the EDSAC.
With most of these programs in the early design phase, Aiken decided to magnanimously spread the knowledge gleaned at Harvard by hosting a conference on “large-scale digital calculating machinery” (computers) alongside the official opening of the Harvard Computation Laboratory. “Aiken felt it was time—there was enough development to get together,” Hopper recalled. “We’d all been isolated during the war, you see, classified contracts and everything under the sun. It was time to get together and exchange information on the state of the art, so that we could all go on from there.”4
Aside from getting together, Aiken saw the Symposium on Large-Scale Digital Calculating Machinery as a forum to highlight the superb work that was being done at Harvard. He believed himself and his loyal crew to be the founders of the field, and he wanted to continue the effort to educate fellow academics as well as industry and government officials about the lessons learned at Harvard. In the words of the January 1947 edition of the Harvard’s alumni bulletin: “They came to see and hear what Professor Howard Hathaway Aiken and his staff had accomplished. To this giant in mathematics the pilgrims paid appropriate homage.”5
Not only would the invitees be impressed with the new facility that housed Mark I, but they would experience a fully operational and automated data-processing center as well. By January 1947 the operation of Mark I had been fine-tuned by Hopper and her staff to the point where the machine ran flawlessly 23 hours in a given day. The machine’s reliability and speed were unmatched by any mechanical or human computation facility, and a significant body of work—including multiple volumes of valuable engineering tables—had been completed. Much of the machine’s output, however, was still couched in military secrecy, so few people were aware of the Computation Laboratory’s accomplishments.6
In typical Aiken fashion, nothing during the symposium would be left to chance. He convened an organizing committee made up of Hopper, Benjamin Moore, and Joseph Harrison.7 The committee identified experts in academia, business, and government who would be interested in the progress of calculating machinery. “Nobody realized it was an industry—it was a branch of research of some kind,” Hopper recalled. “There was beginning to be more people; it was time to pull them together.”8
The committee invited computing leaders from MIT, the University of Pennsylvania, and Princeton to mix with executives from General Electric, Prudential Insurance, National Cash Register, and IBM. The Navy Brass from the Bureau of Ordnance and the Office of Naval Research sat side by side with members of the National Research Council and the National Bureau of Standards. In true fashion, Aiken reserved final judgment on who attended and who spoke. “I made up my mind who was going to be on the program, and I called him up on the telephone and told him he was going to be on the program and what he was going to speak about,” Aiken recalled. “Now I didn’t put words in his mouth, but I told him the subject about which I wanted him to talk.”9
BABBAGE IN THE FLESH
One of the initial people Aiken contacted was Richard Babbage, the great-grandson of the famed mathematician and failed computer-maker Charles Babbage. Aiken first contacted Babbage on 17 April 1946 to inform him about Hopper’s completed book concerning Mark I: “We have recently delivered to the Harvard University Press a manuscript dealing with large-scale calculating machines. This is a book of some six hundred pages, of which the first chapter titled ‘Historical Introduction’ deals in large part with Babbage’s work.”10
In the ensuing correspondence, Aiken discussed Richard Babbage’s great-grandfather’s role in the evolution of calculating machines as described by Hopper; he also inquired about any memorabilia or other information of historical interest concerning his ancestor that Babbage might have in his possession. Aiken explained that the material would be included in an exhibit to be set up in the new Computation Laboratory. This material would be reunited with a piece of Babbage’s Difference Engine that Richard’s grandfather, Admiral H. P. Babbage, had donated to Harvard in 1886.
Richard Babbage was thrilled that a Harvard professor had taken such interest in his obscure and sometimes unappreciated ancestor, and was willing to help Aiken in any way he could. Although the Analytical Engine plans and other mathematical memoirs were part of a display in the Science Museum of London, Babbage did his best to track down other valuable memorabilia within the family. Over the next year, Aiken was supplied with a stream of Charles Babbage’s books, papers, essays, treaties, and tables.
The donations culminated with a rare engraving of the famous mathematician made by Colghin of London in 1832 while Babbage was Lucasian Professor at Cambridge University, the same position held by Sir Isaac Newton. The younger Babbage offered the engraving to Aiken “with the thought that it might hang in your office in proximity to the other mementos and works which you already have.” He believed Aiken to be the intellectual successor of Charles Babbage, and thanked the Harvard professor for providing him with the “unique opportunity to secure for my ancestor a measure of appreciation that did not come his way in life.” The “unique opportunity” of which Richard Babbage spoke was his role as a guest speaker at the Harvard Symposium on Large-Scale Digital Calculating Machinery.11 During his speech Babbage praised Aiken and his staff for persevering in their pursuit for automated computing. This was a quest in which his ancestor fell short, partly because, in his opinion, “all Englishmen [inherently] object to inventions.”12
THE HARVARD SYMPOSIUM
On the morning of 7 January, the first day of the Symposium, the distinguished guests were treated to the first public demonstration of the Computation Laboratory’s second creation, the Harvard Mark II. Unlike Mark I, Mark II was designed and constructed entirely by the Harvard crew. Aiken and Campbell were the chief designers, with periodic input from Hopper and Bloch. “The early thinking about Mark II,” Campbell recalled, “was that it would have relay components, almost completely, both for control and for computations. This would be a departure then from the Mark I, which used electromechanical counters for simulation and for storage.”13
Relays had proven reliable in Mark I, and Aiken decided to extend the technology as a solution for the problem of memory. The crew devised an ingenious latching relay that could stay in a given position without the continuous application of power. The more advanced relay system provided 200 words of memory, while an improved logical design made Mark II twelve times as fast as its ancestor. “The big step forward in Mark II,” Hopper recalled, “was the tremendous development of the multiprocessor and the multi-programmer.”14 Mark II was a dual processor, meaning the machine could be split into two halves that could run independently, or combined in parallel into one multiprocessor. The reason behind this innovation was that one half of the machine could be run against the other as a way to check output accuracy.15
Symposium guests were then escorted from the Gordon McKay Laboratory that housed Mark II to the new Computation Laboratory. Passing the exhibit dedicated to Charles Babbage, the visitors were led to a specially designed gallery with windows two stories high overlooking the Mark I operation. The efficient performance by the machine and its operators was the culmination of years of machine debugging, organizational design, and personnel training. Aiken liked to compare Mark I’s performance to that of a power generating plant. Instead of generating electricity, Mark I and its crew of operators generated numbers.16
The Symposium formally commenced with opening remarks from the administrative vice president of Harvard University, Edward Reynolds. Harvard President James Conant had taken ill and could not attend, perhaps an indication of lingering animosity between Aiken and the Harvard administration. Along these same lines, Reynolds seemed to go out of his way to point out that the developments in comput
ing at Harvard could not have been accomplished without help from IBM, and that the credit for the Automatic Sequence Controlled Calculator must be shared with IBM engineers Clair Lake, Frank Hamilton, and Benjamin Durfee.17
Conversely, Rear Admiral C. Turner Joy, Commanding Officer of the Naval Proving Grounds at Dahlgren, Virginia, highlighted Aiken’s forceful leadership in the new field and applauded the novel relations between Harvard and the Navy. He noted that the modern tools of warfare were now the domain of scientists like Aiken, and was relieved that he personally was too old to have to learn about these new machines. More importantly to many in the audience, Joy pledged that Navy calculators would be made available to all scientists in the future after the backlog of work in guided missiles and aerodynamic supersonic problems had been disposed of. His most audacious statement, however, concerned the future impact of computing machines, for he believed that “the new interest in mathematics might appear more likely to usher in an age of reason than the new interest in atomic energy.”18
Aiken’s comments mirrored the many lectures and presentations he had given on automatic calculating machines since the end of the war. He began with Hopper’s computing history, introducing the audience to “the names of some of the greatest mathematicians and physicists of all time associated with the development of calculating machinery,”19 which led, of course, to his office doorstep at Harvard. These remarks were amplified by the words of Richard Babbage, who later admitted in a letter to Aiken, “I could not help thinking how pleased Charles Babbage would be if he could come alive now and meet the man who discovered how to build a machine that would produce accurate answers to questions which only a genius can put to it!”20