US Military's Research Arm Turns 50
The Advanced Research Projects Agency, basic research, and the Internet
Email Article  Print Article Ronda Hauben (netizen2)    
Sputnik Gives Birth to Important New Research Advances

On Oct. 4, 1957, the world was greeted with a surprise. There was beeping from a man- made object orbiting the earth. This was Sputnik, a 184-pound object the size of a basketball that was to be the catalyst for important new changes in our world. One of these changes would be a significant new means of communications connecting people and computers around the world.

How a small satellite orbiting our globe on Oct. 4, 1957, would, 50 years later, make possible the digitized information and communications network we call the Internet, is a significant story. The subject of this story is, however, not the Internet itself. The subject of the story is the United States research agency that made it possible to create the Internet and other significant computer science developments. This research agency, the Advanced Research Projects Agency, or ARPA as it is more commonly known, was born 50 years ago in February 1958.

This anniversary celebration is a fitting time to look back to how the ARPA began and to ask what this history can teach us about the nature of the kind of research the ARPA was created to support and about the institutional form needed to support such research. Since it can be argued that important achievements of ARPA-supported research include the Internet of today, and other significant computer science advances, understanding the origins and development of the ARPA can set a foundation to understand the origins of the Internet and other computer advances of the past 50 years.

Some Background - The Birth of the ARPA

It is generally recognized that the creation of the ARPA was a direct response to the launch of the world's first orbiting space satellite by the Soviet Union. This was a significant part of the US government's response to the Soviet's surprise achievement. But the mandate of the ARPA was not restricted to space research. A Department of Defense (DOD) directive, number 5105.15, dated Feb. 7, 1958, established "an agency for the direction and performance of certain advanced research or development projects" (1). For reasons to be explained shortly, the director of the agency was to report directly to the secretary of defense. Congressional authorization followed as part of a bill enacted by the US Congress on Feb. 12, 1957.

The Original Mandate

While the ARPA was originally created to support space related research, this function was soon moved to a civilian agency so that space research would have no apparent military connection. The ARPA was thus left to support more general-purpose research.

James Killian, who became the president of the Massachusetts Institute of Technology (1948-1959), and the special assistant for science and technology to President Dwight D. Eisenhower (1957-1959), is credited with establishing the environment in which the ARPA was conceived. Killian had testified at several congressional hearings in the period before Sputnik, advocating for the importance of basic research for the DOD. At those hearings, he and others argued that it was critical to have research that would explore unknown areas in order that the DOD not fall behind in the military and basic research areas of its competition with the Soviet Union. Killian believed that new weapons and weapon systems would require a different form of organization from the traditional roles and missions that the DOD was accustomed to.

Killian described how the great technological successes of the US in World War II such as radar, the proximity fuse and the creation of nuclear weapons were due to how the scientific and technical community functioned even during the war. He drew attention to "the free-wheeling methods of outstanding academic scientists and engineers who had always been free of any inhibiting regimentation and organization." "Every great research laboratory," Killian proposed, "must strive to have men of this kind and to provide an environment analogous to that of the educational institution if it is to be really creative."

Killian believed that the new approaches and weapons systems could not be spawned by the military services themselves. Instead, they could only be expected to "originate in the creative basic research that takes place in the universities and other institutions where fundamental new ideas are most likely to be generated."

Killian argued to Congress that what was needed was research that would be directed toward new concepts and new principles, rather than toward producing pieces of military hardware. He describes why creating an environment to support basic research is of critical importance to the military. "It is" he said, "the yet unanticipated, not yet conceived discoveries which may determine our military strength tomorrow, and we must provide the environment from which such discoveries are most likely to come."

Killian turned the usual argument about basic research and its relevance to the military on its head. Instead of arguing to support research with military objectives, he was arguing for the support for fundamental scientific research because otherwise there would be no possible breakthroughs that could provide relevant research. Unless the DOD provided support for such generalized research, Killian proposed it would fall hopelessly behind its Soviet rival. Similarly, the prestige, which came with being seen as preeminent in science and technology, was critical for the US to maintain its standing in the world.

Articulating this viewpoint explicitly, Killian explained, "The future of the United States, to an extraordinary degree, is in the hands of those who probe the mysteries of the atom, the cell and the stars. Especially is this true of that tiny part of our creative effort which we inadequately term basic research."

Before Sputnik, Killian and his colleagues who argued with him for the primacy for the military of basic research had not been able to have their advice taken seriously. The launch of Sputnik transformed this situation fundamentally.

A report written in 1975 to analyze the ARPA's successes, known as the Barber Report after its main author Richard Barber, depicted the ARPA as having been "spawned in an environment where basic research was equated with military security." Research of a general nature was argued to be the "wellspring" for the advanced ideas critical in the long run for the military.

As the Barber Report explains, this was the changed environment in which the US president at the time, Dwight Eisenhower, supported the creation of the ARPA. Just after the launch of Sputnik, Killian was asked by Eisenhower to recommend how the centrality of basic research could be implemented. Killian recommended the creation of an agency that would support "centers of excellence," flexible funding and long-term, stable environments for researchers. It would be a place where failures were to be seen as expected, to be learned from, and not, as problems.

This was the vision inspiring the creation of the ARPA. Fortunately, in the field of computer science, this vision found champions and the result was that the computer research at the ARPA succeeded in revolutionizing the way that computers would be used in the world.

The Politics of the ARPA

Part of Eisenhower's motive for supporting the creation of the ARPA and its orientation toward basic research, however, had another rationale. This had to do with the problem of rivalry between the different branches of the military services. Eisenhower was opposed to this rivalry, but the DOD having been created only 10 years earlier, in 1947, was still relatively weak in terms of its control over the three different branches of the services. The creation of the ARPA could help to centralize the research done by the DOD.

The services competed vigorously with each other in a number of areas, such as for funding and assignment of new projects. As a result, the creation and placement of the ARPA in the DOD administrative hierarchy became a source of contention between the services and the secretary of defense.

Similarly, since the results of applied research would affect the future of each of the branches of the services, the plan to put applied research in the ARPA met with opposition. In recognition of this political nature of applied research, the secretary of the Air Force, James H. Douglas, said that he was prepared to concede the ARPA a role in basic research but "once you move over the poorly defined line to applied research, I would object." Such pressures defined the environment in which the ARPA began and developed in its early years (2).

Computer Science Is Nourished by the ARPA

Despite these obstacles, the computer science research begun at the ARPA in 1962 was a significant fulfillment of the objectives set out by Killian as the vision for the new agency. In order to understand the ARPA's operations, it is helpful to look at the role played by the director. There have been several different directors in the course of the ARPA's existence.

The period from 1961-63 when Jack Ruina was the director is cited as a particularly formative period. "The Ruina era's legacy," the Barber Report explains, "was particularly important with regard to the ARPA style. It set the precedent of a civilian scientists-director and was characterized by delegation of considerable independence to the technical officers, recruitment of strong technical office directors, minimization of bureaucratic functions and limitation of central program management controls, and stress on quality of staff and contractors."

During the 31-month period that Ruina was director of the ARPA, the computer science program was launched. Computer science was assigned to the ARPA as an area for research in June 1961. The program was originally called Command and Control Research (CCR). The objective of this research was to "provide a better understanding of organizational, informational and man-machine relationships and research on information processing techniques and methods, and maintenance of a general purpose computer facility."

Since in 1961 this was all a new area of research, the services did not have established programs and there were thus fewer constraints on the creation and development of computer science. Ruina soon recruited J. C. R. Licklider, a highly regarded researcher with expertise in psychoacoustics, who had done considerable research on human-machine interaction and computer modeling of the brain's perception of sound. Licklider believed that advances in command and control aspects of computing would require fundamental advances in the field of computer science. He was particularly interested in developing the area of interactive computing (3).

Ruina gave Licklider a free hand to create a computer science research program. Just as Killian would have advised, Licklider began by creating a set of "centers of excellence" at several universities, each of which would focus on a particular area of computing research. He changed the emphasis, which had been on command operational studies, war game scenarios and command system laboratories, to research in time-sharing systems and interactive computing, computer graphics, improved computer languages and computer networking.

By early 1964, the name of the computer science research office at the ARPA was changed to the Information Processing Techniques Office (IPTO), to reflect the changes in the research program Licklider had introduced. Among the centers of excellence the IPTO set up were one at MIT, known as Project MAC, and one at Carnegie Mellon. Licklider writes that one center was to "lead the effort to achieve balance in information technology, to harness the logical powers of computers to make it truly available and useful to men." The other was to "lead the effort to achieve fundamental understanding to develop the theoretical bases of information processing" (4). Subsequently, other centers of excellence were set up, including one focusing on computer graphics.

Though computer networking was part of Licklider's plan for the research to develop the computer science field, during his first two-year period at the ARPA, it was too early for this area of research. The program initiated by Licklider in computer science led to the ARPA being recognized throughout the field, according to the Barber Report, "as being the main supporter and perhaps the most important force in the course of the US and probably world history in the computer."

The goal of Licklider's program in computer science was to develop the computer in ways other than number crunching. This led to what became perhaps the most significant area of computer development at the IPTO. This involved the recognition that the computer could be a communication device, which led to the research developing packet switching and the ARPANET, and subsequently, the research creating TCP/IP and the Internet.

Describing the paradigm change represented by computer networking research, Michael Hauben writes:

Fundamental to the ARPANET, as explained by the [ARPANET] Completion Report, was the discovery of a new way of looking at computers. The developers of the ARPANET viewed the computer as a communications device rather than only as an arithmetic device. This new view made building the ARPANET possible. This view came from the research conducted by those in academic computer science. Such a shift in understanding the role of the computer is fundamental to advancing computer science. The ARPANET research has provided a rich legacy for the further advancement of computer science and it is important that the significant lessons be learned and studied and used to further advance the study of computer science. (5)
This perspective shift in how to view the computer, especially in looking at the computer as a communication device, was the basis for the area of research that represents probably the greatest achievement of the IPTO and of the ARPA.

This is the area of research first developing the ARPANET and subsequently providing the practical and conceptual leadership for the creation and spread of the Internet (6).

The ARPA and the Struggle Within

Critical to an understanding of the ARPA, however, is the understanding that the struggle both within the agency itself and in the creation and support for the agency was a continual battle between the objectives and practices of the military and the objectives and practices of the researchers who were working for the IPTO or in its programs. By the 1970s, the researchers at the IPTO were subjected to serious constraints.

A directive issued on March 23, 1972, by the DOD replaced the ARPA's 1959 charter with a new one. The name of the ARPA was changed to the Defense Advanced Research Projects Agency (DARPA). This removed the agency from its original position within the Office of the Secretary of Defense. The administrative placement of the agency was changed from where it had been placed to protect it from the competition of the services. At the time there was a concern that the separation of the ARPA from the Office of the Secretary of Defense would weaken it and its independence.

Describing the significance of moving the ARPA from the protection of the Office of the Secretary of Defense, Charles Herzfeld, director of the ARPA from 1965-67, writes:

But one fundamental change to DARPA is more important than all these vicissitudes. In 1958, the body was designed to be an agent for change in the Department of Defense, located in the Office of the Secretary of Defense. In the 1960s, it became stronger and more effective in this role. Sometime in the 1970s or '80s, the agency shrank to being an agent for change in the Office of the Under Secretary of Defense for Acquisition, Technology and Logistics, which focuses on building and buying weapons. (7)
Licklider, too, was disturbed by the changes that occurred at the ARPA when he returned as director of the IPTO in January 1974. He found that much had changed. He observed that, "there was really much less opportunity to initiate things.… At that time [the ARPA director] had a fixed idea that a proposal is not a proposal unless it's got milestones. I think that he believed that the more milestones, the better the proposal.… Milestones had to be written into the proposal and it was completely rewritten" (8).

In an e-mail message to IPTO researchers in April 1975, Licklider writes:

[There is a] development in ARPA that concerns me greatly -- and will, I think, also concern you. It is the continued and accelerating (as I perceive it) tendency on the part of the ARPA front office, to devalue basic research and the effort to build an advanced science/technology base in favor of applied research and development aimed at directly solving on an ad hoc basis some of the pressing problems of the DOD. (9)
The Barber Report notes again the importance of the organizational placement of the agency if the agency is to be able to support basic research. "During its first decade, ARPA's leadership tended to feel that the agency was a unique organization in DOD with special ties to the secretary [of defense] and hence somehow immune from the impact of many forces and decisions that shape the activities of the [military] services and other parts of the department."

By the post 1967 period, this protected position was changing, so that the ARPA was more constrained than it had been previously.

The authors of the Barber Report are not surprised by the changes, but they are struck by how little attention is paid to them and "the relative lack of discussion or debate" among the leadership of the Department of Defense.

With the celebration of the 50th anniversary of the ARPA, there is renewed attention being paid to reviewing the experience of this agency. Such a review of the experience of the ARPA is pregnant with the lessons of the importance of government support for basic research.

The past 50 years provides a set of achievements demonstrating the importance of the initial vision that Killian and other scientists in the 1950s advocated regarding the importance of basic research.

These voices, however, were ignored until Sputnik was launched. Only then did the necessity for the federal support for basic research become inescapable. The ARPA and its initial orientation toward supporting basic research is the product of these events.

The organizational structure of the ARPA made possible the creation of the computer science research office within the ARPA begun by Licklider. That office has demonstrated the importance of the support for basic research in the field of computer science. The IPTO supported a general area of research, one with a far-reaching impact. The achievements of this research office were not specific defense-related applications, nor were the goals narrowly aimed at defense specific applications. If this reality is not recognized, however, it is possible to attribute significant computer science achievements mistakenly to defense specific objectives.

A common and widespread myth exists that the Internet has grown out of a defense specific objective, that is, from the goal to create a computer network that could survive a nuclear war. This is a striking example of how a false narrative can spread and gain public credence.

This false narrative finds its roots in the failure to understand that the ARPA was not an agency created for defense specific applications, but to support the basic research that would lead to new concepts and ideas.

Only then could the new conceptual frameworks become available in general and, in that context, also for defense related developments. If one starts with the goal of creating defense specific developments, however, the research is limited and not able to go beyond what is known at the time.

In summing up this relationship between the ARPA, the IPTO and basic research, Alan Perlis, one of the IPTO researchers explains: "We owe a great deal to ARPA for not circumscribing the directions that people took in those days. I like to believe that the purpose of the military is to support ARPA and the purpose of ARPA is to support research" (9).
1. The Barber Report says that the secretary of defense actually issued the directive creating the ARPA on Feb. 4, 1957. Unless otherwise indicated quotes are from the report. The report is online.

2. Barber Report, p. I-27.

3. This was a period when computer use generally required that the programmer bring a program typed on punch cards to a computer facility and return several hours later to get a print out of the program's results. This form of computing was known as batch processing.

4. Ronda Hauben, "Computer Science and the Role of Government in Creating the Internet, Part III Centers of Excellence and Creating Resource Sharing Network."

5. Michael Hauben, " Behind the Net: The Untold History of the ARPANET and Computer Science," in "Netizens: On the History and Impact of Usenet and the Internet."

6. Ronda Hauben, "The Internet: On Its International Origins and Collaborative Vision (A Work in Progress)."

7. Charles Herzfeld, "How the Change Agent Has Changed," Nature, Vol. 451, Jan. 24, 2008, p. 404.

8. Thomas Bartee, ed. "Expert Systems and Artificial Intelligence," Indianapolis 1988, p. 225. See Ronda Hauben, "Computer Science and the Role of Government in Creating the Internet: ARPA/IPTO (1962-1986) Creating the Needed Interface."

9. Adele Goldberg, "The History of Personal Workstations," ACM, New York 1988, p. 129. See also Ronda Hauben, "The Birth and Development of the ARPANET" in Michael Hauben and Ronda Hauben, "Netizens: On the History and Impact of Usenet and the Internet." There is also a print edition published by John Wiley and Sons, 1997.

--

This article was written for Futurezone and appears in German on its Web site. Futurezone is the Technology web site for Orf, Austria's national public broadcast media.

2008/02/12 오후 1:13
© 2019 Ohmynews
◀ Return to Article