by Kurt W Beyer
THE CONTENTS OF HOPPER’S MANUAL
The manual began with a historical summary of the development of mechanical aids from the abacus up to Aiken’s idea for Mark I. Hopper’s history outlined the trajectory of technological progress and created a sense of connection between the handful of talented individuals who advanced the computing art. Hopper’s history is interesting on a variety of levels. First, it is one of the earliest attempts to document the historical progression of the field. Second, and more importantly, it provides insight into Hopper’s view of technological change.
Because of travel limitations during the war, Hopper’s historical research was limited to resources found in Harvard University’s Widener Library.36 The first significant figure introduced to the reader by Hopper is Blaise Pascal. Pascal (1623–1662) is highly regarded as a Christian thinker and as a scientist-mathematician, but few acknowledge his skills as an engineer. Pascal the engineer was responsible for inventions as diverse as the medical syringe, the hydraulic press, the wheelbarrow, and a horse-drawn public transportation system used in Paris during the 1660s. Pascal’s father, Etienne, served as the chief tax collector for the Rouen district in France during the 1640s, a job that required a vast amount of tedious calculations. To escape the drudgery of his father’s work, the clever youth devised plans for a machine that would automate the process and produce error-free results. The basic machine was a metal box containing a system of wheels and cylinders that could add, subtract, and carry over numbers. The most difficult problem Pascal faced was how to carry over numbers automatically. Teaching a child to add 14 plus 8 and carry over a 1 to the tens column to arrive at 22 is simple, but to teach a machine to do the same is another matter. Pascal invented an ingenious weighted ratchet system connected to the counters in each wheel. With each turn of the dial to the next notch, the ratchet was raised. When the respective wheel passed 9, the ratchet sprung and tripped a weight that turned the next wheel to the left one space and the original wheel back to 0. The direct actuation of a numbered wheel and its carrying method, Hopper wrote, “are the foundation on which nearly all mechanical calculating machines since have been constructed.”37
The next significant individual on Hopper’s historical path leading to Aiken and Mark I was Gottfried Wilhelm von Leibniz (1646–1716). Leibniz, the consummate genius of his day, excelled as a mathematician, a diplomat, a lawyer, a historian, and a philosopher. Apart from his more celebrated accomplishments, Leibniz was also responsible for an extraordinary automated calculating machine, which he completed in 1694. Hopper commended Leibniz for achieving multiplication and division through an ingenious system of stepped wheels. These wheels, similar to the inner workings of a mechanical clock, were interconnected in such a way that mathematical solutions rather than the time of day were their output. According to Hopper, the device’s “stepped reckoner” design was so mechanically eloquent that it would be included in most calculating machines thereafter and serve as an important link to Mark I.38
The most significant section of the manual is dedicated to the English philosopher-mathematician Charles Babbage. The story of Babbage’s computing engines begins with his honeymoon in 1819. While touring Europe, Charles and his new wife met the prominent French civil engineer Gaspard Francois de Prony. Years earlier, while making tables for the fledgling French Republic, de Prony had applied an idea he had encountered while reading Adam Smith’s Wealth of Nations (1776). Inspired by Smith’s concept of the division of labor, de Prony applied it to his teams of mathematicians. One group considered formulas, another applied them numerically, and the third did the computations. The efficiencies created by such a division produced more accurate tables in less time. Babbage borrowed this concept and made it a part of his machine designs.
Between 1820 and 1822, Babbage developed the plans for his first computing machine, which he named the Difference Engine. The Difference Engine would calculate a table for any linear mathematical function by using Newton’s method of differences, which allowed complex polynomial equations to be solved without any need to multiply or divide. The first prototype of the Difference Engine required a variety of precise manufacturing processes and quickly evolved into the most expensive government-funded research project of its day. But by the mid 1830s the expensive project began to unravel for lack of concrete results. To make matters worse, Babbage began championing a completely new design concept. His novel idea developed into the blueprints for his Analytical Engine, a device that broke from earlier calculating machines. But his change of course cost Babbage his supporters in government and scuttled any chance of his finishing the Difference Engine.
The Analytical Engine was not just an extension of the Difference Engine; it was a new concept in automated information processing. Information would be continuously fed into the “mill” of the machine by means of the ingenious punchcard system that Joseph Jacquard used to automate his textile looms. Hopper emphasized that two distinct decks of cards fed the machine: “The first set was designed to select the particular numbers to be operated upon from the store; the second set, to select the operation to be performed by the mill. Since the deck of operational cards represented mathematical solutions independent of the variables involved, the analytical engine was general in regard to algebraic operations.”39
The one person who truly understood the ramifications of Babbage’s new machine was a beautiful and eccentric woman 23 years his junior. Ada Byron King, Countess of Lovelace and daughter of the renowned poet Lord Byron, was a mathematician who had become interested in Babbage’s work. In 1834 she began to write regularly to the older mathematician, and the two became close friends and colleagues. By 1840, Babbage had completed the basic blueprints for the Analytical Engine. Ada Lovelace wrote seven technical essays that provided powerful explanations for what the machine could do and how it could do it, including what today would be called a computer program.
Lovelace’s work, published in 1843, did not aid Babbage in successfully securing the vast amounts of funding needed to construct the Analytical Engine. The aging mathematician had used up his goodwill with the government on the uncompleted Difference Engine. Babbage would not be able to turn his idea into reality during his lifetime. According to Hopper, the world had to wait until 1943 before Babbage’s dream would materialize in the laboratory of Howard Hathaway Aiken.
THE HOPPER VERSION OF HISTORY
Grace Hopper’s account of computing’s early history is interesting on a variety of levels. It is the story of a series of brilliant men separated by vast expanses of time but connected by the dream of automating the drudgery of mathematical calculations. Each one was personally responsible for introducing radical and potent conceptual ideas, which were then constrained by the technologies of the day. Their advances were discontinuous and independent from the work of others. Such a historical interpretation sheds light on the name Aiken gave to his creation: Mark I. Mark I was a one-of-a-kind, discontinuous invention.
The historical account also provides the reader a sense that with the publication of the Mark I manual the computational torch had been officially passed to Howard Aiken and his team. The epigraph to chapter 1 of the manual includes this passage from Babbage’s autobiography:
If, unwarned by my example, any man shall undertake and shall succeed in really constructing an engine embodying in itself the whole of the executive department of mathematical analysis upon different principles or by simpler mechanical means, I have no fear of leaving my reputation in his charge, for he alone will be fully able to appreciate the nature of my efforts and the value of their results.40
In a very real sense, Aiken believed that Babbage was speaking directly to him through the years. Shortly after the publication of the Manual, Aiken developed a personal relationship with Babbage’s great-grandson and decorated his office with Babbage memorabilia, including books once owned by Babbage with marginal notes in Babbage’s own hand. He even acquired (through the Harvard library) original p
arts of the partially constructed Difference Engine.41
When the Harvard crew moved into the new Computation Laboratory facilities in 1946, the Babbage memorabilia were gallantly exhibited in the vestibule. For members of the laboratory crew, Aiken’s copy of Babbage’s autobiography, Passages from the Life of a Philosopher,42 became mandatory reading. Hopper recalled being handed a copy when arriving at Harvard in the summer of 1944. “The book [Babbage’s Autobiography] gave you a feeling that the development was inevitable,” she remembered. “It was bound to come.” Passages from the Life of a Philosopher also exposed Hopper to Ada King for the first time: “She wrote the first loop. I will never forget; none of us ever will.” Nor did people let Hopper forget the uncanny coincidence of history, casting Aiken in the role of Babbage and Hopper in the role of Lady Lovelace.43
Commander Aiken and Lieutenant (j.g.) Hopper posing with a piece of Charles Babbage’s difference engine for the Christian Science Monitor, March 1946. Courtesy of Archives Center, National Museum of American History, Smithsonian Institution.
THE IBM AUTOMATIC SEQUENCE CONTROLLED CALCULATOR: SAME MACHINE, DIFFERENT HISTORY
About 6 months after Hopper was assigned the task of writing the Mark I Manual, an anonymous technical writer at IBM produced a pamphlet aptly named “IBM Automatic Sequence Controlled Calculator.” This competing document provided a general description of the same machine, yet its differences speak volumes about underlying conceptions. Like Hopper’s manual, the first section of the pamphlet summarized the history leading up to the machine and its creation:
The IBM Automatic Sequence Controlled Calculator dedicated at Harvard University on Aug. 7, 1944 marked the latest advance in International Business Machines Corporation’s program of adapting its equipment to use in the field of scientific computation, in which for many years it has been collaborating with leading universities and research organizations.44
One can quickly glean the direction of the IBM narrative from the opening sentence of the document. First, the assertion that the dedication marked the “latest advance” suggests a continuous process in which the Automatic Sequence Controlled Calculator was just the newest iteration. The “corporation’s program of adapting its equipment” further emphasizes continuity of process, and reminds the reader that the technology used in this machine was no different than that already employed in other IBM business calculating machines. Finally, the very idea of applying business machines for scientific computation is not novel, as IBM “has been collaborating with leading universities and research organizations” for many years. That collaboration, according to the IBM narrative, began in 1928 with the founding of the Columbia University Statistical Bureau. The Bureau was established solely “for the purpose of assisting other educational institutions and scientific and research organizations in adapting IBM mechanisms to their computing requirements.”45 It was here that machines such as the IBM Difference Tabulator and the IBM Summary Card Punch were developed. In 1934 a separate laboratory was established in the Columbia Department of Astronomy, and since 1937 the Thomas J. Watson Astronomical Computing Bureau, a joint enterprise of IBM with the American Astronomical Society and Columbia University, operated the laboratory.
Because IBM had been applying standard machinery to scientific calculations since the 1920s, it was no wonder that “Harvard University’s need for a machine such as the IBM Automatic Sequence Controlled Calculator had long been a topic of discussion among members of the Harvard Faculty.”46 Howard Aiken’s inquiry had been but “one of these discussions.” Moreover, it was Professor Harlow Shapley (director of the Harvard Observatory) and T. H. Brown (Professor of Business Statistics) who pointed out to Howard Aiken that “IBM standard equipment for some time had been successfully used for scientific calculation purposes.”47
The IBM narrative squarely attempts to marginalize Aiken and his role in the development of the machine. It was Professor Brown “who brought the subject to the attention of Mr. J. W. Bryce, dean of IBM’s scientists and inventors,” and it was Bryce who determined “the company’s ability and willingness to build the machine.”48 Acting on Brown’s and Bryce’s recommendations, Aiken met with IBM engineers and “outlined the University’s requirements.”49 The IBM history does not acknowledge Aiken as the initiator of the project or as the inventor of the machine. Furthermore, it reminds the reader that Aiken was called to active duty in the Navy and thus spent a significant time away from the project.
In fact, the IBM document shrewdly underlines Aiken’s limited role with Aiken’s own words, quoting strategically from his laudatory speech at the machine’s dedication ceremony on 7 August 194450:
We approached the International Business Machines Corporation and asked their support to build such a machine and construct it and put it into operation Our first contact with that company was with Mr. J.W. Bryce. Mr. Bryce for more than thirty years has been an inventor of calculating machine parts, and when I first met him he had to his credit over 400 fundamental inventions—something more than one a month. They involved counters, multiplying and dividing apparatus, and all of the other machines and parts which I have not the time to mention, which have become components of the Automatic Sequence Controlled Calculator that you are to see this afternoon.51
With this vast experience in the field of calculating machinery, our suggestion for a scientific machine was quickly taken and quickly developed. Mr. Bryce at once recognized the possibilities. He at once fostered and encouraged this project, and the multiplying and dividing unit included in the machine is designed by him.
On Mr. Bryce’s recommendation, the construction and design of the machine were placed in the hands of Mr. C.D. Lake, at Endicott, and Mr. Lake called into the job Mr. Frank E. Hamilton and Mr. Benjamin M. Durfee, two of his associates.52
Indeed, Aiken, in his own words, appears to give credit for the invention, design, construction, and operation of the machine to IBM engineers.
DIFFERENT HISTORIES, DIFFERENT AGENDAS
Why do the historical accounts written by Grace Hopper and IBM differ to such a great extent? One obvious explanation attributes the divergence to the well-documented animosity between Thomas Watson Sr. and Howard Aiken.53 On the surface this is reasonable, for evidence can be found that self-importance drove each man to call the project his own. But in this case, it seems that history is being employed as a tool for ambitions greater than individual pride.
For Thomas Watson Sr., IBM interests were best served by replacing individual history with organizational history. The locus of technological innovation, according to IBM, was the corporation. The myth of the lone radical inventor working in the laboratory or basement was replaced by the reality of teams of faceless organizational engineers contributing incremental advancements in the name of the company.54 To drive this point home, the IBM narrative includes the following list:
Of the many basic units of the Calculator, invented or developed by IBM engineers, the more important are:
• Multiplying Machine, invented by Bryce in 1934.
• Dividing Machine, invented by Bryce in 1936.
• Multiplying Dividing Machine invented by Bryce and Dickinson in 1937.
• Unit Counter invented by Carroll in 1925.
• Ratchet Type Plate Counter, invented by Lake and Pfaff in 1935.
• Pluggable Type Relay, invented by Lake and Pfaff in 1937.
• Double-Deck Card Feed, invented by Lake in 1921.
• Electromatic Typewriter developed for automatic operation by Lake and Hamilton in 1936.
• Counter Readout and Emitter, invented by Bryce in 1928.
• Commutator Total-Taking Mechanism, invented by Daly in 1926.55
The majority of these technical advances predate Howard Aiken’s inspiration for Mark I, serving as further proof of his marginal role. Not even Aiken’s inspiration, Charles Babbage, was safe from IBM’s historical scrutiny. Again, applying Aiken’s own words , “I say Babbage failed but I wou
ld like to make it especially clear that he failed because he lacked the machine tools, electrical circuits, and metal alloys”56 In other words, Charles Babbage failed because he lacked the support of a corporation such as IBM.
IBM hoped to direct the future of calculating machines by controlling the past. The Automatic Sequence Controlled Calculator was not a disruptive technology that would spawn a new calculating industry separate from IBM’s comparative advantage in punch-card equipment.57 Rather, innovation would be contained within the research facilities of Endicott and the patents retained by the organization. Eventually, IBM planned to dominate the scientific and mathematical markets, much as it controlled the business market for punch card machines.
For Aiken and Hopper, on the other hand, the connection with notable geniuses such as Babbage served practical agendas both inside and outside the Computation Laboratory. Years later, Hopper stated: “I’m quite sure that he [Aiken] discovered Babbage well after he had the concept of the engine and that he used it as a selling thing that made it more legitimate.”58 In fact, I. Bernard Cohen has considered the question of how much Babbage’s machines influenced Aiken, precisely because Aiken had equated himself with Babbage as far back as 1937.59 It appears however, that Aiken did not covet Babbage’s ideas so much as he coveted his prestige. To be associated with the man who held the same chair of mathematics at Cambridge as Sir Isaac Newton would help the rather undistiguished 37-year-old graduate student turn diagrams and notes into an 81-foot machine that weighed 9,445 pounds, contained 530 miles of wire, and cost $750,000.60
With the completion of the Mark I manual, the need for legitimacy, as well as the benefits from it, had grown exponentially. The general demobilization at the end of World War II affected research projects at universities across the United States. Federal plans for a return to normalcy necessitated termination of wartime contracts, which meant that Commander Aiken and his machine would have to find sustainable sources of revenue to keep the operation a float. This was particularly daunting since relations with IBM were in shambles. Moreover, Aiken’s position at Harvard was somewhat precarious, given his lack of tenure, a volatile disposition, and the administration’s tepid support for applied mathematics.
