Jean Jennings Bartik: Pioneer of Electronic Computer Programming
On February 1, 1946, reporters were given the opportunity to observe the first public trial of a new device that promised to do the work of a year in a week. It was called ENIAC, the world’s first electronic computer, and the somewhat befuddled newsmen had it explained to them as the brainchild of John Mauchly (1907-1980) and J. Presper Eckert (1919-1995), but what they weren’t told was that the only reason it was functioning at all for them on that day was due to the programming efforts of the six women who had ushered them in and taken their coats, and were now standing silently in the back, unacknowledged throughout the whole groundbreaking presentation.
The ENIAC Six had been given no guidance on how to go about the task of programming this new machine, because there was no guidance to give. Electronic programming did not, as of yet, exist. All the team had were the schematics for ENIAC and its eighteen thousand vacuum tubes, and a directive to make it all work, somehow, while the engineers built it. Each one of those women would become a hero of early computing history, and in time we shall tell all of their stories, but today is reserved for the tale of Jean Jennings Bartik (1924-2011), the small town softball ace who became one of the era’s most sought after visionaries in the field of general use computing.
She was born on December 27, 1924 in Gentry County, Missouri, which at the time boasted a population of some 15,000 souls. The county had been shrinking since 1900, and would continue to do so throughout Jennings’s life and on into the present day, and Jean’s little corner of it was a town called Alanthus Grove, population 104 in 1925, most of whom were related to Jean in one way or another. They all attended the same one-room school together when they weren’t working on the farm, and Jean, the sixth of seven siblings, was able to skip fifth grade by listening to the older students’ lessons and snooping around her older siblings’ work. She was a field-plowing, cow-milking farmgirl who also happened to be something of a genius in mathematics, in a family that revered education as much as it loved working the earth, and when she was ready for high school at the age of twelve she was sent to live in Stanberry (population 2,029 at the time) because Alanthus Grove was too far off the bus route.
Jean shone in high school. She devoured math and science classes, led the student council, edited the high school section of the town paper, and drew massive (by Stanberry standards) crowds whenever she took to the pitcher’s mound for the girls’ softball team, dominating visiting squads and one time hurling a no-hitter that is probably still talked about to this day. She graduated in 1941 as the salutatorian of her class, and made her way, as so many in her family had before her, to Northwest Missouri State Teachers College at the age of sixteen.
She did not have long to enjoy her college idyll, however, as war came to the United States in December of 1941, and the campus emptied of its male students and faculty, gone to fight the Axis powers. Jean continued with her mathematical studies as the campus re-stocked itself with Navy sailors taking the courses they needed for the war effort, one of whom she dated until he was called overseas and they made the joint decision not to attempt to keep the relationship alive once he left.
The summer of her junior year, the financial help she received from her family ran out, and she worked as a silver-plater at Kansas City’s Pratt & Whitney plant, plating the metal onto propeller gears. The work was generally wretched, breathing in degreasing fumes, working with silver cyanide vats, and lacerating one’s hands on steel wool, but thousands of women leapt into jobs like these during the war out of a feeling of patriotism and the desire to do their part to support their husbands, sons, and fathers overseas. Jennings never viewed her time there quite that romantically, and worse still, the money would only cover one more semester of school, but fortunately the Math Department at NMSTC brought in retired professors to teach her everything she had left to learn within the space of a single semester. It was a tremendous push, but she made it, receiving her degree just in time to take advantage of one of history’s most unique opportunities.
Beginning in 1942, the University of Pennsylvania’s Moore School had begun actively recruiting women skilled in mathematics for work at the Ballistics Research Laboratory (BRL). The problem of artillery has always been to get your guns on target before the enemy does so with theirs, so if you can figure out, for a certain type of gun, using a certain type of ammunition, in a certain type of weather, what angle will cause your shells to land a certain location without a bunch of just random guessing and firing, that would lead to significantly improved results in war. The women hired at BRL were doing just that, using a mechanical desk calculator to churn, line by line, and moment by moment, through the path of an artillery shell under a given set of circumstances, creating a complete trajectory in the matter of a couple of days of brute force computation.
This wasn’t mindless work - it required an understanding of the differential equations that governed the paths of these objects at every moment of their flight, and the first call for recruits was for women who had already earned a mathematics degree, though later the constant shifting of the theater of war (and with it shifting of environmental factors) required an up-scaling of the number of women performing calculations, who didn’t necessarily need degrees, and would instead receive their training in higher mathematics on site before getting their trajectory assignments.
It was repetitive work, but it was directly contributing to the war effort, while also using mathematical skills professionally in a way that most women who received math degrees a generation before never had the opportunity to, and the floors of the BRL calculating buildings soon filled up with promising young math minds. The most accomplished of these were selected to work with the differential analyzer, a sprawling analog computer in the basement that could be set up to solve complicated trajectory problems quickly, but that had to be essentially rebuilt for each new problem posed to it, with snapped wires and bands constantly slowing down the process. What was needed was something that could do the work of the differential analyzer but without the physical limitations of its component parts.
Enter ENIAC. Mauchly had long been thinking about the potentials for electronic computation, in spite of skeptics who said flatly that the number of vacuum tubes required rendered such a device a practical impossibility. Tubes were considered to be highly unreliable, and the idea of something depending on hundreds or thousands of them working correctly in concert was laughable. Mauchly brought the idea to Eckert, a young whiz kid engineer, who pondered over the problem, and judged it surmountable. In 1943 they received the go-ahead from BRL to start construction on their 18,000-tube dream machine, and while they were busy building it from scratch, the question arose, once it is built, how do we break problems up so that ENIAC can solve them?
This world, of dedicated women churning through computations one crank at a time while elsewhere, in secret, a team was giving birth to the age of electronic computation, was the one Jean found when she arrived in Philadelphia on March 30, 1945, having responded to one of BRL’s far-flung calls for women mathematicians, instantly hopping on a train headed East the moment she received word of her acceptance. She began work, like everybody else, computing trajectories step by step with a hand calculator, but in June a memo from on high described an opportunity to work on a secret new device being constructed.
Jean went to the interview, where her candidacy was supported by Adele Goldstine, the elegant and inspirational computing instructor who had taught Jean when she first arrived. Jean, who had only been working at BRL for a couple of months, did not make the first cut, but was chosen as a second alternate, meaning that when two of the original selections had to bow out, the job was offered to her, and she responded with an enthusiastic Yes.
Jean and the other four women chosen for the project travelled to Aberdeen Proving Ground in mid-June, where they were put through a crash course on the IBM punch card reader system, making their way through the material on an army base of some twenty-seven-thousand men. One month later, they were back in Philadelphia, armed with a full knowledge of the IBM punch machine, and ready to get started on their mystery project. They would have to wait a month, but finally, in August, a mass of technical papers in long rolls was brought before them. These were the schematics of ENIAC. The group was not yet allowed to see the physical device, and were instructed to start developing the methods by which mathematical problems could be loaded onto the device, just from the schematics alone.
Nothing like this had ever been done before, and the women received effectively no guidance in these early stages for where to begin, or how to proceed. They had the schematics, their knowledge of IBM punch cards, their experience with trajectory computations, and their mathematical knowledge, and that was it. They soon decided to divide the mammoth project up into three chunks, with two people working on each major component of ENIAC (a sixth member had since been added to the team, bringing the Eniac 6 up to its full and permanent strength). Jean worked with Betty Snyder to understand the functionality of the accumulator, deducing that it could not only be used to do arithmetic, but that it could store ten decimal numbers, to become a temporary storing house for numbers that would be needed in later operations. After putting their discoveries together with those made by the other two teams, the ENIAC 6 had a pretty good idea of how the machine worked, and a few tantalizing ideas about how aspects of its functionality could be exploited to make computations run more smoothly, as when Kay McNulty realized that the structure of ENIAC could allow the results of a given calculation to be fed back in as the inputs for the next round of that calculation, creating what we think of now as a Loop.
Jean and Betty’s next big project was, essentially, inventing the basics of modern programming. Jean took the larger scale mathematics and figured out ways to represent it as a cascade of smaller order computations, and Betty figured out ways to represent those computations as operations in ENIAC. The key here was coordination - some processes ran faster than others, and so some results had to be held while other parts of the machine caught up, so that they could both be fed into the next step simultaneously. They ordered their thinking through the creation of “pedaling sheets” - massive pieces of paper with 27 columns across the top, each representing a different primary component of ENIAC, with each row below representing a step of the program they wanted to run, allowing them to graph out visually how each process flowed through the machine so that intermediate values got to the unit they needed to get to at precisely the moment when that unit was ready to receive it.
Looking through the pedaling sheets, Betty and Jean realized that, if they wanted to solve trajectory calculations as fast as possible and use as much of ENIAC at any given time as possible, they would have to stop thinking serially, in terms of one long stream of calculations that went from beginning to end, and start thinking in terms of parallel computing. What processes could be started alongside each other, initiated by the same program pulse, to ensure that as little of the machine was lying idle at any moment as possible, and everything still came together when it needed to in order to compute the trajectory? Out of their own inner need for absolute efficiency, Jean and Betty produced the first electronic parallel programs.
Betty, for her part, invented the concept of regular internal check-points which would bring a calculation to a stop if a certain condition weren’t met, which would serve the team well in the long process of directly programming ENIAC when they at long last got direct access to it in November of 1945. Their foundational work proved so solid that it was a matter of only a couple of months before their program was ready for its first public demonstration which inaugurated the age of electronic computation.
Unfortunately, as we saw above, Jean was not given any public acknowledgment during this first demonstration of the program she had helped bring to life, and the ceremony proved something of a dreadful anti-climax to her months of hard and brilliant work, but the people involved with the project knew her value, and when it came time to make the next big leap in programming complexity, she was appointed as a central element of that push. She started working with Richard Clippinger on ways to get ENIAC to run his wind tunnel turbulence problems, but they soon ran into a problem - the equations governing the turbulent phenomena Clippinger was studying were just too big to run on the ENIAC as it existed, so Jean started working on a suggestion made by John von Neumann (1903-1957) to use the ENIAC’s storage capacities to store instructional code instead of just numerical values. This retooling of ENIAC into a “stored code” computer required a fundamental rethink of how programs were written and wired, and the lead work for this throughout 1947 fell to Jean, and once again she succeeded in creating a functioning new approach to programming where before there was nothing.
During this time, heavy-handed administration had pushed Mauchly and Eckert away from Moore School, into their own business venture creating the world’s first commercial computers, which they duly invited Jean to join. This would lead to the creation of BIVAC and UNIVAC I, and to the reunion of Jean with Betty Snyder (now Betty Holberton) to produce the first program created by a computer program in 1951. She would go on to work at Remington Rand when it bought Eckert-Mauchly out, but the birth of her first child in 1954 led her to retire from the world of programming until her divorce gave her a second wind. She received her Master’s Degree in 1967, and eventually assumed the editorship of the Auerbach Minicomputer Reports in the 1970s, using the publication to educate companies about how the new revolutions in computers could aid their businesses, opening a whole new segment of the world to the miraculous advances in problem solving she had helped usher into existence a quarter century before.
Jean continued working in the computer field until 1986, when she switched to a job in real estate, where she remained employed for the next quarter century, staying alive long enough to witness the resurgence of interest in her pioneering computing work, including books, documentaries, and museums celebrating what she and the other ENIAC 6 programmers had achieved, even publishing her own memoir of those years shortly before her death in 2011, something almost unheard of for women who had made their scientific mark in the 1940s and 50s. Even if she hadn’t, however, the magnitude of her contribution to our modern world is so great, its permeation into every aspect of our societal interactions so profound, that we would never have forgotten her, and that phenomenal mind which created order from nothingness.
FURTHER READING:
Kathy Kleiman’s wonderful Proving Ground (2022) is the first place to go for the stories of the ENIAC 6. The writing is brisk, the explanations are accessible to anyone, and Kleiman did work we can’t thank her enough for in gathering first-hand accounts from the Six about their experiences in the 1940s and 50s. The next place to go would be Jean’s memoir, Pioneer Programmer: Jean Jennings Bartik and the Computer that Changed the World but goooooood luck finding one, friend. Every six months or so I try and flag down a copy for myself, and haven’t yet, but perhaps fortune will smile upon you!
Comments