ACM Established to Develop Communication About Computing

In 1987, Mina Rees noted that 40 years earlier "there was a relatively small group that played a central role" in computer development [8]. Even with their small numbers, though, computer scientists found it difficult to stay abreast of the changes and advances in their rapidly changing field. In January 1947, Howard Aiken from the Harvard Computational Laboratory organized a symposium to gather people who were interested in sharing their activities and knowledge regarding mathematical and computational advances for calculating machinery. Aiken estimated about 60 people would attend this four-day symposium; 336 actually did attend. The field was growing off scale.

In his opening address to the symposium, Aiken remarked [1]:

"A great variety of new techniques has been introduced into the design of calculating machinery….There has been inadequate publications of results, and inadequate transmittal of results from one group working in the field to another. Consequently, we have often found that we were working on problems that had already been solved by others. We have found that we were beginning researches that were nearing completion in other laboratories. Those are precisely the reasons why this symposium seems so necessary at this time."

The importance of robust communication around computer development was underscored at the symposium by Rear Admiral Charles Turner Joy, whose introduction of Aiken noted: "…military and peacetime uses [of mathematics] are interlinked…Perhaps it is not beyond the realm of fantasy to predict that future developments in the art of building mechanical brains may produce one capable of contributing to the analysis of the problems of international relationships" [7]. Mina Rees, chair of the Mathematical Methods session at the 1947 symposium, summed up the situation: "Around the county there was clear evidence of an increasing desire to stay in touch with the rapidly changing technological scene and to establish improved means of communication among the many persons who were professionally interested in the new machines" [8].

In the minds of mathematicians involved in post-war computer development, establishing a communication system to "stay in touch" had important national security and quality-of-life implications. Conversely, if computer developers did not stay in touch, the U.S. could face unnecessary risks from Cold War adversaries and its way of life could be threatened.

In his closing address to the symposium, MIT electrical engineering professor Samuel Caldwell restated the importance of establishing a communication system for computer developers and mathematicians [5]. He identified two impediments to establishing an open information-sharing system: private industry’s patent concerns and the military’s concerns for classifying information sensitive to national security. Caldwell reviewed the arguments supporting both of these concerns and their consequences for computer development. Traditionally, communication about technology development was compartmentalized and controlled either within private companies or within divisions of military organizations. Yet, as Aiken noted at the beginning of the symposium, such controlled communication practices resulted in inefficient technology development, since researchers often unknowingly duplicated their efforts when they could not communicate across organizational barriers.

Caldwell ended his presentation by calling for the establishment of a professional association to support systematic communication among computer developers: "It is this communication system that I have described as consisting of two elements: an organization to make us more immediately and more continuously aware of a common purpose and thus furnish the incentive for communication; and a medium for such communication" [5].

Mina Rees remembered that "when Sam Caldwell stood before the audience attending the Symposium on Large Scale Calculating Machinery at Harvard and called for the establishment of a new association to provide for better communication among those interested in the new machinery, it struck a responsive note" [8]. But not everyone in the computer field was so eager to start yet another association and publication. Edmund Berkeley recounted the reactions he received to this idea: "John von Neumann said, `Oh, for heaven’s sake, not another association.’ And George Stibitz said, `Oh, not another association.’ …But I remarked to several of the people there that if the prominent people would not have an association, the second-level people might very well have an association. And so we put together a meeting in September at Columbia University, …I believe something like 57 people attended, and one of the resolutions passed was to form an Eastern Association for Computing Machinery" [4].

Rees recounted that "Ed Berkeley [from Prudential Insurance] and John Curtiss [from the Bureau of Standards] were particularly active in urging the establishment of the new EACM. John became its first president, and Ed, the first secretary" [1]. Franz Alt later recounted, "During this early period, and for some years thereafter, the burden of work for the association fell mainly upon E.C. Berkeley" [2]. Alt and Rees joined this initial group as members of the Executive Committee.

Back to Top

Establishing Communication and a Profession

At the first meeting of the EACM, a resolution was passed setting out the organization’s purpose: "…to advance the science, development, construction, and application of the new machinery for computing, reasoning, and other handling of information. Its territory is the Eastern states. By removal of Eastern from the name, the organization may be converted into a national organization if and when that becomes desirable" [6]. At this meeting, Berkeley presented a letter he received from von Neumann that was cautiously supportive of the new association [6]:

"I am much interested in the work and development of your organization. As you may know, my personal opinion was, and still is, that such an association is a highly desirable one ultimately but that the general situation has not yet matured sufficiently to make the present moment the optimum one to found it. This does not mean, however, that I will not be very glad if you succeed in furnishing the proof of the opposite, and I want to use this occasion to wish you the best of luck in your efforts."

At that September 1947 meeting, 223 people representing 74 organizations had expressed interest in participating in the EACM. Two months later, "Eastern" was dropped from the name and the Association for Computing Machinery became the first national (and later international) society of computing professionals, now poised to celebrate its 60^th anniversary [4].

The first ACM conference was held in December 1947 at the Aberdeen Proving Ground Ballistic Research Laboratories with over 300 people attending. There were 13 papers presented, covering the cutting-edge topics of that time [3]:

• General Principles of Coding, with Application to the Eniac (John von Neumann)
• Adaptation of the Eniac to von Neumann’s Coding Techniques (R.F. Clippinger)
• Optimum Size of Automatic Computers (George Stibitz)
• The Univac (John Mauchly)
• The Raytheon Computer (R.V.D. Campbell)
• Operating Characteristics of the Aberdeen Machine (Franz Alt)
• Reduction of Doppler Observations (D. Hoffleit)
• On the Approximate Solution of a Partial Differential Equation on the Differential Analyzer (Joseph Levin)
• Non-Linear Parabolic Equations (Bruce Hicks)
• Computation of the Airflow about a Cone Cylinder (M. Lotkin)
• Air Flow Problem Planned for the Eniac (R.F. Clippinger)
• Laminary Boundary Flow in a Compressible Fluid (John Holberton)
• Census Applications of High-Speed Computing Machines (James McPherson)

Summaries of these papers were distributed to ACM members through the mail with meeting reports and through publication in the National Research Council’s quarterly Mathematical Tables and Other Aids to Computation [3]. This conference and the wide distribution of information presented there represented a tangible solution to the communication problem facing computer developers at the end of World War II.

ACM publications have been an important tool for achieving the association’s mission since its inception. Edmund Berkeley was painstaking in maintaining a list of computer professionals, which he initially distributed under the title "A Roster of Organizations in the Computing Machinery Field" [5]. The work of identifying people in the computing field and establishing a common technical vocabulary for their work may seem trivial to 21st century computer developers. Sixty years ago, however, these activities were the foundation for establishing computer science as a professional field in its own right, separate from electrical engineering and mathematics. Mina Rees related a contemporaneous ONR project that worked in the same way to establish the computing profession [8]: "The first task [at the ONR] resulted in the publication of a book that reported on the status of developments of computing-machine components in 1950. This task was an outgrowth of my early discussions with Tompkins about the state of the computing art and was based on a conviction we shared—that a significant contribution to the development of the new computers and their integration into the many fields of application would be served by consistent attention to a broad dissemination of information about advances being made."

These compilations of the knowledge in a field traditionally serve to define and delimit a developing profession. By identifying members in the budding profession, creating opportunities for these professionals to interact, recording and disseminating knowledge through regular publications, and partnering with allied associations, the ACM created a nexus around which the fledgling computer industry could come together to collaborate and support collective, interdisciplinary efforts for national security and the betterment of our quality of life.

This article is available at doi.acm.org/10.1145/1230819.1230834.

Back to Top

Back to Top