Under pressure because of the USSR’s continuous rocket launches, the Republican President set up ARPA despite considerable Congressional and military dissent. Although it scaled back some of its original goals, ARPA went on to subsidize the creation of computer science departments throughout the country, funded the Internet, and consistently supported projects that enhanced human/computer interactivity.

Forming ARPA

Headquartered in the Pentagon, ARPA was established to develop the US lead in science and technology innovations applicable to the military and help it respond quickly to any new challenges. Eisenhower had multiple suspicions about the military and its industrial connections. However, he did believe in basic research and appointed a man with similar notions, Neil McElroy, the head of Proctor & Gamble, as his Secretary of Defense. McElroy pushed his vision of a “single manager” for all military-related research through Congress. Despite objections by the heads of the various armed forces, Eisenhower sent a request to Congress on January 7, 1958, for startup funds to create ARPA and appointed its director, a vice-president from General Electric. Shortly after, Congress appropriated funds for ARPA as a line item in an Air Force appropriations bill.[1]

Roy Johnson came to head ARPA from GE, dreaming of human-crewed space stations, military moon bases, orbital weapons systems, global surveillance satellites, and geostationary communications satellites. But by the end of ARPA’s first year, Eisenhower had established NASA, dashing Johnson’s space fantasies. Space projects moved to the new civilian agency or back to the individual military services, including the covert ones like those of the CIA’s spy planes and satellites. ARPA desperately searched for a new mission and argued effectively for going into “basic research” areas that were considered too “far out” for the other services and agencies.

With the Kennedy Administration taking office and its appeal for the nation’s “best and brightest” to enter government service, ARPA found its prospects improving. It looked aggressively for talent to develop the best new technologies. Behavioral research, command and control, missile defense, and nuclear test detection were some of the newest projects taken on by ARPA, although not necessarily “basic” research. The new agency also got increasingly involved with computers, especially after Joseph Carl Robnett “JCR” Licklider joined the staff in October 1962.[2]

ARPA’s Information Processing Techniques Office (IPTO)

The IPTO emerged in the early 1960s with the charge of supporting the nation’s advanced computing and networking projects. Initially called the Office of Command and Control Research, its mandate was to extend the knowledge gained by researching and developing the multi-billion dollar SAGE (Semi-Automatic Ground Environment) project and extend it to other command and control systems for the military.

SAGE was a joint project by MIT and IBM with the military to computerize and network the nation’s air defense system. It linked a wide array of radar and other sensing equipment throughout Canada and the US to what was to become the Colorado-based NORAD headquarters. SAGE was meant to detect aircraft (bombers and later ICBMs) coming over the Artic to drop nuclear bombs on Canada and the US. The “semi-automatic” in SAGE meant that humans would be a crucial component of the air defense system, and that provided an opening for Licklider’s ideas.

SAGE consisted of some 50 computer systems located throughout North America. Although each was a 250-ton monster, SAGE computers had many innovations that further sparked the dream of man-machine interactivity. These included data communications over telephone lines, cathode ray terminals to display incoming data, and light pens to pinpoint potential hostile aircraft on the screen. ARPA’s IPTO helped transform SAGE innovations into the modern IT environment.

From Batch to Timesharing

Throughout the 1960s, three directors at IPTO poured millions of dollars into projects that created the field of computer science and got computers “talking” to people and to each other. Licklider had the Office of Command and Control Research changed to Information Processing Techniques Office (IPTO) when he moved from BBN to ARPA to become its first director. Licklider was also from MIT, but what made him unusual was that he was a psychologist amongst a majority of engineers. He got his Ph.D. from the University of Rochester in 1942 and lectured at Harvard University before working with the Air Force. Foremost on his agenda was to encourage the transition from “batch processing” to a new system called “timesharing” to promote a more real-time experience with computers, or at least a delay measured in seconds rather than hours or days.

These new developments meant the opportunity for new directions, and Licklider would provide the guidance and government’s cash. During the mid-1950s, Licklider worked on the SAGE project focusing mainly on the “human-factors design of radar console displays.”[3] From 1959 to 1962, he was a Vice-President for BBN, overseeing engineering, information systems, and psycho-acoustics projects. He was also involved in one of the first time-sharing cases at BBN with a DEC PDP-1 before taking a leave of absence to join ARPA for a year.[4]

Licklider swiftly moved IPTO’s agenda towards increasing the interactivity of computers by stressing Vannevar Bush’s ideas and the notion of a more personal and interactive computing experience. An influential military project at MIT was the TX-2, one of the first computers to be built with transistors and a predecessor to the PDP line of computers. It also had a graphics display, unlike most computers that used punch cards or a teletypewriter. The TX-2 was located at MIT’s Lincoln Laboratories and had a major influence on Licklider. The brilliant psychologist would ride the waves of Cold War grant monies and champion research and development for man-machine interactivity, including a radical new computer-communications technology called timesharing.

Early computer users submitted their requests and punch cards to a receptionist at a computer center. Then a team of computer operators would run several (or a “batch”) of these programs at a time. The results were usually picked up a day or two after submitting the requests. After Bell Labs developed transistor technology, individual transistors were wired into circuit boards, creating the “second generation” computer series. This new technology allowed vacuum tubes to be replaced by a smaller, cheaper, and more reliable technology and produced an exciting increase in processing speeds. Faster technology eventually led to machines that could handle several different computing jobs at one time – timesharing.

Time-sharing would allow several users to use a computer by taking advantage of the increasing processing speeds. It also used enhanced computer communications by allowing users to connect via teletype and later cathode-ray terminals. Rather than punching out programs on stacks of paper cards and submitting them for eventual processing, time-sharing made computing a more personal experience by making it more immediately interactive. Users could interact with a large mainframe computer via teletypewriters used originally for telex communications and the cathode-ray terminals used for televisions.

Timesharing emerged from the MIT environment and its support by the US government. Sets of procedures used for timesharing originated at MIT after receiving an IBM 704 in 1957, a version of the AN/FSQ-7 developed for SAGE. John McCarthy, a Sloan Fellow from Dartmouth, recognized some possibilities of sharing the computer’s capabilities among several users. As the keyboard replaced punch cards and magnetic-tape-to-magnetic-tape communication as the primary source of data entry, it became easier for the new computers to switch their attention to various users.[5]

As its human users paused to think or look up new information, the computer could handle the requests of other users. Licklider pressed the notion of timesharing to increase the machine’s interactivity with humans, but the rather grandiose vision would not be immediately accepted throughout the military-related sphere of ARPA. It was still in a relatively primitive state of computing in the early 1960s, but ARPA would soon be won over.

First on Licklider’s list was Systems Development Corporation (SDC), a RAND spin-off that had done most of the programming for the SAGE project. ARPA had inherited SDC, and a major part of the IPTO budget was set to help them transition from the SAGE air defense project to command and control computing. SDC had been given one of SAGE’s ANSFQ-32 mainframes, but to Licklider’s chagrin, they used it for batch processing. Licklider thought it ridiculous to use it in this manner, where responses often took hours or even days to help a commander react to battle situations.[6] Licklider immediately went to work to persuade SDC to switch from batch processing to time-sharing, including bringing in his allied colleagues such as Marvin Minsky for seminars to cajole SDC.

Soon they were convinced, and Licklider moved on to other time-sharing projects, pouring ARPA money into like-minded projects at MIT and Carnegie Mellon. Luckily, he had joined ARPA the same month as the Cuban Missile Crisis. The event raised concerns about the ability of the President and others high on the chain of command to get effective information. In fact, Kennedy had been pushing for better command and control support in the budget, reflecting his concerns about being the Commander-in-Chief of a major nuclear power.

In the next part I will examine timesharing and the first attempts to commercialize it as a utility.

Notes

[1] Background on ARPA from Hafner, K. and Lyon, M. (1998) Where Wizards Stay Up Late. New York: Touchstone. pp. 20-27.
[2] A much more detailed version of these events can be found in a chapter called “The Fastest Million Dollars,” in Hafner, K. and Lyon, M. (1998) Where Wizards Stay Up Late. New York: Touchstone. pp. 11-42.
[3] Information on Licklider’s involvement with SAGE from Campbell-Kelly, M. and Aspray, W. (1996) Computer: A History of the Information Machine. Basic Books, pp. 212-213.
[4] Information on JCR Licklider’s background at BBN from the (2002) Computing Encyclopedia Volume 5: People. Smart Computing Reference Series.
[5] Evans, B.O. “Computers and Communications” in Dertouzos, M.L. and Moses, J.(1979) The Computer Age: A Twenty Year View. Cambridge, Massachusetts: The MIT Press. p. 344.
[6] A good investigative job on Licklider and SDC was done by Waldrop, M. Mitchell (2001) The Dream Machine: J.C.R. Licklider and the Revolution that Made Computing Personal. New York: The Penguin Group.

Ⓒ ALL RIGHTS RESERVED

---

Anthony J. Pennings, PhD is Professor and Associate Chair of the Department of Technology and Society, State University of New York, Korea. Before joining SUNY, he taught at Hannam University in South Korea and from 2002-2012 was on the faculty of New York University. Previously, he taught at St. Edwards University in Austin, Texas, Marist College in New York, and Victoria University in New Zealand. He has also spent time as a Fellow at the East-West Center in Honolulu, Hawaii.

Related

Category: how IT came to rule the world, science and technology studies, the cold war
Tags: AN/FSQ-7 > ANSFQ-32 > ARPA > batch processing > J.C.R. Licklider > SAGE (Semi-Automatic Ground Environment) > Systems Development Corporation (SDC)

Comments

Comments are closed.