Grace Hopper and Information-Age Invention

By Colin E. Babb

Hopper Google Doodle

Computer pioneer Grace Hopper is famous for what she created—Google honored her achievements in 2013—but she also should be remembered for how she did it.

The terms “Information Age” and “Computer Age” are often conflated, as if both describe adequately the technological, social, economic, and political forces that have driven human societies over the past 70 years toward an interconnected world in which information itself has become a commodity. When the first modern computers were being developed in the 1940s, however, it was not at all clear just what kind of information these new machines ultimately would be able to create. Electromechanical computers such as the Harvard Mark I and even the more sophisticated ENIAC were capable only of doing what Mark I designer and operator Howard Aiken described as “makin’ numbers.” They were essentially elaborate calculators, the creations of mathematicians to be used for mathematical problems. Had the inputs and outputs of computers remained solely a matter of numbers, it is doubtful that the “Computer Age” would have led as quickly to the “Information Age” that we live in today. A naval officer, Grace Hopper, helped make that transformation a reality by employing a distinctive method of technology development.

For most of the history of computing, which has seen computers do everything from regulate refrigerators to help put men on the moon, the fundamental language of every computer remained simply zeroes and ones (quantum computing—where bits can be both 0 and 1 at the same time—is changing this, but that is another story). Every program is at its lowest level merely a sequence of “off” and “on” commands. It was apparent to some in the first generation of programmers in the 1940s and 1950s that programming in this fashion—writing (or punching) an endless series of zeroes and ones—would be, at best, mind-numbing and, at worst, a constant invitation to the incorporation of errors for anyone but the most expert programmer. The challenge for these earliest programmers was how to create a system that would help humans— who don’t naturally communicate in mathematics—to write programs in reasonably understandable script that could be “translated” into the numbers-based language of computers. The problem of compatibility operates in two directions: It is important to facilitate the work of programmers who create inputs in the form of programs, and it is essential to ensure that outputs are understandable to someone who cannot read binary or some other code.

Grace Hopper influenced the decisive first part of this equation by playing a key role in developing one of the first widely used programming languages, COBOL (Common Business-Oriented Language). Her career with computers took her from academia to the military to businesses large and small and back again—a matrix of destinations that would become familiar to many of those who entered the field after her. She often is portrayed as a pioneer for women in mathematics (few women took doctorates in the field when she received hers in 1934) as well as computing. Yet her role as a pioneer transcends gender, and, as Kurt Beyer argues in his Grace Hopper and the Invention of the Information Age (2009), a key to her success was her collaborative style of invention (what he terms “distributed invention”).

Hopper began her professional life teaching mathematics at Vassar College in Poughkeepsie, N.Y., in the early 1930s. An important part of her experience there, according to Beyer, was her eagerness to take advantage of the faculty’s ability to audit classes in other departments. In taking everything from physics to astronomy to economics, Hopper broadened her own intellectual world while also incorporating what she learned back into her teaching. While the experience rubbed some of her colleagues the wrong way—by her own recollection, it was the younger faculty who took the most umbrage at her not concentrating solely on mathematics—it prepared her for working with a wide array of partners in her subsequent career in computing. It also helped her to see the ways in which mathematics could contribute to understanding better other fields and vice versa.

With the outbreak of World War II, mathematics would be enlisted in the service of every major combatant in ways barely contemplated in previous wars. From building new ships, aircraft, and armored vehicles to creating new weapons and even splitting the atom, mastering the sheer avalanche of numbers necessary for planning and fighting the war became a Herculean task. The first modern computers were developed to tackle these and other complex problems in multiple places, all within a few years of each other, in Germany, the United Kingdom, and the United States. In the latter, the first computers were tasked with solving ballistics problems and creating firing tables for the Army, with its ENIAC (under construction during the war but not completed before its end), and the Navy, with Harvard University’s Mark I. The Mark I also would be used for problems associated with the Manhattan Project, but even the computer’s operators would be unaware during the war that they were providing help to this top secret endeavor. It was here, too, that Hopper would enter the world of computing in the summer of 1944 as a newly commissioned lieutenant junior grade in the Naval Reserve, assigned to a small team of operators of the new “calculator.”

GRACE HOPPER’S CAREER IN THE NAVAL RESERVE SPANNED MORE THAN 40 YEARS. SHE RETIRED IN 1986 WITH THE RANK OF REAR ADMIRAL (LOWER HALF) AT THE AGE OF 79. AT THE TIME, SHE WAS THE OLDEST ACTIVE-DUTY OFFICER IN THE NAVY. (PHOTO COURTESY OF HARVARD UNIVERSITY ARCHIVES)

GRACE HOPPER’S CAREER IN THE NAVAL RESERVE SPANNED MORE THAN 40 YEARS. SHE RETIRED IN 1986 WITH THE RANK OF REAR ADMIRAL (LOWER HALF) AT THE AGE OF 79. AT THE TIME, SHE WAS THE OLDEST ACTIVE-DUTY OFFICER IN THE NAVY. (PHOTO COURTESY OF HARVARD UNIVERSITY ARCHIVES)

The Mark I, designed by professor and naval reservist Howard Aiken and built by IBM, was more than 50 feet long and composed of more than 750,000 parts. A spiritual, if not physical, descendent of Charles Babbage’s 19th-century difference engine, the electromechanical Mark I used punched paper tape for its instructions and output. Hopper, who quickly became adept at working with the huge machine, earned Aiken’s confidence, such that she soon was asked to write the Mark I’s operating manual. This was the first of what would become a long line of publications penned by Hopper that would both document her own education as a computer evangelist and define the nascent field for its first generation of users and programmers. Remaining with the Harvard team even after the end of the war, Hopper became a manager of a data-processing center that operated 24 hours a day, seven days a week.

In 1949, Hopper joined what is widely regarded as the first computer “startup,” the Eckert-Mauchly Computer Corporation (EMCC). Run by two of the creators of the ENIAC, J. Presper Eckert and John Mauchly, EMCC sought to build a commercial version of that computer—what would eventually be called the UNIVAC. Although the company’s independent existence would be short-lived (it would be purchased in 1950 by Remington Rand after a series of financial difficulties), Hopper would find the experience there memorable. An important relationship there, biographer Kathleen Broome Williams points out, was a friendship with Betty Holberton, whom Hopper believed was the first person to use a computer to write a program.

While working at Remington Rand (later Sperry Rand), Hopper began to confront the formidable and interconnected problems of the time-consuming nature of programming and a still-low number of people entering the highly specialized field of computing. Automating part or all of the programming process by saving or compiling groups of code used over and over by different programs offered a way to reduce the time spent on writing programs and, eventually, to help non-math specialists engage with computers. The first attempts at building these compilers (beginning with A-0 in 1951) put Hopper on the forefront of computing in the early 1950s, but they were difficult to use and still required significant amounts of programming time. In May 1954 at a conference sponsored by the Office of Naval Research (ONR), Charlie Adams, a researcher with MIT’s Project Whirlwind, presented the work of several other MIT programmers. Their “algebraic compiler” took standard math symbols and translated them into computer language. Two years later, at a second ONR-sponsored conference, Hopper acknowledged that this work was the most comprehensive language at the time. Beyer argues that MIT’s work helped Hopper go in a new direction with her compilers, leading her to believe that the twin goals of opening up who could be a programmer and ending the disconnect with users were within the field’s grasp.

By the beginning of 1959, both a widespread need for “automatic” programming as well as the tools required to build it were in place. In April of that year, Hopper met with Robert Bemer of IBM and Howard Bromberg of RCA to discuss the possibility of creating a common business language. The small group found a willing sponsor and partner in Charles Phillips, director of data systems research at the Department of Defense. The group that would create COBOL came to be called the Conference on Data Systems Languages (CODASYL). The group’s first meeting in May 1959 consisted of representatives from the Air Force’s Air Material Command, the Commerce Department’s Bureau of Standards, the Navy’s David Taylor Model Basin, Honeywell, Burroughs Corporation, IBM, RCA, Sylvania, and Sperry Rand. As CODASYL’s work continued through the 1960s, this list of both industrial and government participants grew greatly in size.

Although Hopper’s is the name most often associated with the COBOL, the language’s invention was “distributed,” which Beyer defines as a style in which prototypes are farmed out to an ever-widening circle of creators. Work on COBOL began with three committees, each with a different task: evaluating current languages (there were others in use, such as FORTRAN), analyzing the syntax of languages, and determining how to build a language useful for both business and science. The group was greatly influenced by one of Hopper’s compilers, FLOW-MATIC, in creating COBOL, which would use English-like words and syntax as the fundamental basis for the whole language. What had until then been an experience of writing code in numbers, equations, and symbols became with COBOL a matter of mastering commands such as “GO TO,” “DISPLAY,” and “STOP RUN.” It was not high literature—but it was an enormous step forward when CODASYL submitted its final report on the project in December 1959.

CMDR. HOWARD AIKEN (LEFT), LT. J.G. GRACE HOPPER, AND ENS. ROBERT CAMPBELL IN FRONT OF THE HARVARD MARK I IN AUGUST 1944, NOT LONG AFTER HOPPER CAME ON TO THE PROJECT. THE JUXTAPOSITION WITH THE SAILORS IN THE BACKGROUND REVEALS THE OPERATING PROCEDURE OF THE MARK I, WHERE ENLISTED PERSONNEL WERE OPERATORS AND OFFICERS WERE PROGRAMMERS. (PHOTO COURTESY OF HARVARD UNIVERSITY ARCHIVES)

CMDR. HOWARD AIKEN (LEFT), LT. J.G. GRACE HOPPER, AND ENS. ROBERT CAMPBELL IN FRONT OF THE HARVARD MARK I IN AUGUST 1944, NOT LONG AFTER HOPPER CAME ON TO THE PROJECT. THE JUXTAPOSITION WITH THE SAILORS IN THE BACKGROUND REVEALS THE OPERATING PROCEDURE OF THE MARK I, WHERE ENLISTED PERSONNEL WERE OPERATORS AND OFFICERS WERE PROGRAMMERS. (PHOTO COURTESY OF HARVARD UNIVERSITY ARCHIVES)

At the time, CODASYL represented a distinctive method of technology invention: using not simply a team, but a team of teams, from multiple organizations. This collaboration of diverse business and government interests brought together potential customers and colleagues—as well as past and future rivals. They all understood, however, that there was an enormous advantage to working together on what promised to be a computer common language that would potentially make everyone’s job easier. Programming could now be a job for anyone, not merely for a select few who understood the fine points of mathematical equations. The creation of COBOL and subsequent languages allowed the 1960s, as Beyer points out, to be the first decade in which programmers could finally concentrate on data processing problems rather than on simply mastering the operation of machines.

Interdisciplinary and interagency collaboration and team-oriented management are now ubiquitous in the science and technology community. CODASYL’s legacy remains with us today as an exemplary method of invention and research—the Internet itself, for instance, would be “invented” using an industry-academia-government collaboration in 1969, and today’s naval science and technology community also remains committed to this kind of approach to managing invention.

The spirit in which CODASYL operated—to provide an open-source code for the benefit of an entire community without regard solely to profit or individual advantage—also remains alive and well. Globally distributed, collaborative projects have built such widely used software programs as Firefox, OpenOffice, and Linux, which are free to all users. Created and maintained by individuals and groups that donate their time and labor, these types of projects are a product of what NYU professor Clay Shirky has called the “cognitive surplus”—the collective potential productive capacity created by modern labor-saving technology that can be applied to social projects. Magnified by the power of global communication networks—themselves enabled by computers—Hopper’s method of distributed invention can now be replicated on a massive scale.

Hopper’s experience with the creation of COBOL in particular, as well as her career trajectory in general, exemplified another major development of the postwar science and technology community: the rise of what President Dwight Eisenhower termed in 1961 the “military-industrial complex.” Indeed, Hopper’s work at Harvard, EMCC, and Remington Rand while a naval reservist not only was at the very intersection of this partnership (which also definitively included academia), it demonstrated that the linkages in many respects were already mature when she arrived on the scene. Hopper did not create these connections—she took advantage of and played a role in strengthening them.

Through Grace Hopper, the Navy had an important role in the invention of the first widely used programming language—but the story of COBOL is a modern tale, where collective projects end up being more than the sum of their parts. Ultimately, the creation of viable computer programming languages redefined not only what computers could do, but also who could operate and use them. How information was created and stored—even the very definition of what constituted knowable information itself—was transformed. Computers were no longer machines built only for “makin’ numbers.”

 About the Author:

Colin Babb is a contractor who serves as the historian for the Office of Naval Research and managing editor of Future Force.

About Future Force Staff