Where Wizards Stay Up Late: The Origins of the Internet

by Katie Hafner

  What time-sharing could not do was eliminate the necessity of coordinating competing demands on the machine by different users. By its nature, time-sharing encouraged users to work as if they had the entire machine at their command, when in fact they had only a fraction of the total computing power. Distribution of costs among a number of users meant that the more users the better. Of course, too many users bogged down the machine, since a high percentage of the machine’s resources were allocated to coordinating the commands of multiple users. As the number of users increased, more of the computer’s resources were dedicated to the coordination function, which reduced actual usable processing time. If programmers had to do very small jobs (such as tightening code or minor debugging of a program), they didn’t need a powerful machine. But when it came time to run the programs in full, many of which used a lot of machine resources, it became apparent that users were still in competition with one another for computing time. As soon as a large program requiring a lot of calculations entered the mix of jobs being done on-line, everyone’s work slowed down.

  • • • When the Air Force passed the Q-32 on to ARPA in 1961, Ruina didn’t have anyone to administer the contract. Ruina had in mind a job with the potential for expansion far beyond the single contract that happened to be pressing at the moment: Computers, as they related to command and control, might one day provide high-speed, reliable information upon which to base critical military decisions. That potential, largely unfulfilled, seemed endlessly promising.

  Coincidentally, Ruina also was looking for someone who could direct a new program in behavioral sciences that DOD wanted ARPA to run. By the fall of 1962, Ruina had found the candidate who could fill both posts, an eminent psychologist named J. C. R. Licklider.

  Licklider was an obvious choice to head a behavioral sciences office, but a psychologist wasn’t an obvious choice to oversee a government office focused on developing leadingedge computer technology. Yet Licklider’s broad, interdisciplinary interests suited him well for the job. Licklider had done some serious dabbling in computers. “He used to tell me how he liked to spend a lot of time at a computer console,” Ruina recalled. “He said he would get hung up on it and become sort of addicted.” Licklider was far more than just a computer enthusiast, however. For several years, he had been touting a radical and visionary notion: that computers weren’t just adding machines. Computers had the potential to act as extensions of the whole human being, as tools that could amplify the range of human intelligence and expand the reach of our analytical powers.

  Joseph Carl Robnett Licklider was born in St. Louis in 1915. An only and much-beloved child, he spent his early years nurturing a fascination with model airplanes. He knew he wanted to be a scientist, but he was unfocused through most of his college days at Washington University. He switched concentrations several times, from chemistry to physics to the fine arts and, finally, to psychology. When he graduated in 1937 he held undergraduate degrees in psychology, mathematics, and physics. For a master’s thesis in psychology, he decided to test the popular slogan “Get more sleep, it’s good for you” on a population of rats. As he approached his Ph.D., Licklider’s interests narrowed toward psychoacoustics, the psychophysiology of the auditory system.

  For his doctoral dissertation, Licklider studied the auditory cortex of cats, and when he moved to Swarthmore College, he worked on the puzzle of sound localization, attempting to analyze the brain’s ability to determine a sound’s distance and direction. If you close your eyes and ask someone to snap his fingers, your brain will tell you approximately where the snap is coming from and how far away it is. The puzzle of sound localization is also illustrated by the “cocktail party” phenomenon: In a crowded room where several conversations are taking place within one’s hearing range, it is possible to isolate whatever conversation one chooses by tuning in certain voices and tuning out the rest.

  In 1942 Licklider went to Cambridge, Massachusetts, to work as a research associate in Harvard University’s Psycho-Acoustic Laboratory. During the war years, he studied the effects of high altitude on speech communication and the effects of static and other noise on reception by radio receivers. Licklider conducted experiments in B-17 and B-24 bombers at 35,000 feet. The aircraft weren’t pressurized, and the temperatures on board were often well below freezing. During one field test, Licklider’s colleague and best friend, Karl Kryter, saw Licklider turn white. Kryter panicked. He turned up the oxygen and yelled to his friend, “Lick! Speak to me!” Just as Kryter was about to ask the pilot to descend, the color returned to Licklider’s face. He had been in tremendous pain, he said, but it had passed. After that, he stopped partaking of his favorite breakfast—Coca-Cola— before going on high-altitude missions.

  By this time, Licklider had joined the Harvard faculty and was gaining recognition as one of the world’s leading theorists on the nature of the auditory nervous system, which he once described as “the product of a superb architect and a sloppy workman.”

  Psychology at Harvard in those years was strongly influenced by the behaviorist B. F. Skinner and others who held that all behavior is learned, that animals are born as blank slates to be enscribed by chance, experience, and conditioning. When Skinner went so far as to put his own child in a so-called Skinner box to test behaviorist theories and other faculty members began doing similar experiments (albeit less radical ones), Louise Licklider put her foot down. No child of hers was going into a box, and her husband agreed.

  Louise was usually the first person to hear her husband’s ideas. Nearly every evening after dinner, he returned to work for a few hours, but when he got home at around 11:00P.M.he usually spent an hour or so telling Louise his latest thoughts. “I grew up on his ideas,” she said, “from when the seeds were first planted, until somehow or other he saw them bear fruit.”

  Everybody adored Licklider and, at his insistence, just about everybody called him “Lick.” His restless, versatile genius gave rise through the years to an eclectic cult of admirers.

  Lick stood just over six feet tall. He had sandy brown hair and large blue eyes. His most pronounced characteristic was his soft, down-home Missouri accent, which belied his acute mind. When he gave talks or led colloquia, he never prepared a speech. Instead, he would get up and make extensive remarks off the cuff about a certain problem he happened to be working on. Lick’s father had been a Baptist minister, and Louise occasionally chided him by noticing the preacher in him. “Lick at play with a problem at a briefing or a colloquium, speaking in that soft hillbilly accent, was atour de force,” recalled Bill McGill, a former colleague. ”He’d speak in this Missouri Ozark twang, and if you walked in off the street, you’d wonder, Who the hell is this hayseed? But if you were working on the same problem, and listened to his formulation, listening to him would be like seeing the glow of dawn.”

  Many of Lick’s colleagues were in awe of his problem-solving ability. He was once described as having the world’s most refined intuition. “He could see the resolution of a technical problem before the rest of us could calculate it,” said McGill. “This made him rather extraordinary.” Lick was not a formalist in any respect and seldom struggled with arcane theorems. “He was like a wide-eyed child going from problem to problem, consumed with curiosity. Almost every day he had some new fillip on a problem we were thinking about.”

  But living with Lick had its frustrations, too. He was humble, many believed, to a fault. He often sat in meetings tossing ideas out for anyone to claim. “If someone stole an idea from him,” Louise recalled, “I’d pound the table and say it’s not fair, and he’d say, ‘It doesn’t matter who gets the credit; it matters that it gets done.’” Throughout the many years he taught, he inspired all his students, even his undergraduates, to feel like junior colleagues. His house was open to them, and students often showed up at the front door with a chapter of a thesis or just a question for him. ”I’d put my thumb up and they’d pound up to his third-floor office,” said Louise.

&nbs
p; In the postwar years, psychology was still a young discipline, inviting derision from those in the harder sciences with little patience for a new field that dealt with such enigmatic entities as the mind, or “the human factor.” But Licklider was a psychologist in the most rigorous sense. As one colleague put it, he belonged with those “whose self-conscious preoccupation with the legitimacy of their scientific activity has made them more toughminded than a good many of their colleagues in the better established fields.”

  By 1950, Lick had moved to MIT to work in the Institute’s Acoustics Laboratory. The following year, when MIT created Lincoln Laboratory as a research lab devoted to air defense, Lick signed on to start the laboratory’s human-engineering group. The cold war had come to dominate virtually the entire intellectual life of the institution. Lincoln Lab was one of the most visible manifestations of MIT’s cold war alliance with Washington.

  In the early 1950s many military theoreticians feared a surprise attack by Soviet bombers carrying nuclear weapons over the North Pole. And just as scientists had coalesced during the 1940s to deal with the possibility of German nuclear armament, a similar team gathered in 1951 at MIT to deal with the perceived Soviet threat. Their study was called Project Charles. Its outcome was a proposal to the Air Force for a research facility devoted to the task of creating technology for defense against aerial attack. Thus Lincoln Laboratory was quickly formed, staffed, and set to work under its first director, the physicist Albert Hill. In 1952, the lab moved off-campus to Lexington, about ten miles west of Cambridge. Its main projects centered around the concept of Distant Early Warning—the DEW line: arrays of radars stretching, ideally, from Hawaii to Alaska, across the Canadian archipelago to Greenland, and finally to Iceland and the British Isles. Problems of communication, control, and analysis for such an extended, complex structure could be handled only by a computer. To satisfy that requirement, Lincoln first took on Whirlwind, a computer project at MIT, and then developed a successor project called the Semi-Automatic Ground Environment, or SAGE.

  Based on a large IBM computer, SAGE was so mammoth that its operators and technicians literally walked inside the machine. The system served three major functions: receiving data from various detection and tracking radars, interpreting data relating to unidentified aircraft, and pointing defensive weapons at incoming hostile aircraft. SAGE was only “semiautomatic,” since the human operator remained an important part of the system. In fact, SAGE was one of the first fully operational, real-time interactive computer systems. Operators communicated with the computer through displays, keyboards, switches, and light guns. Users could request information from the computer and receive an answer within a few seconds. New information continuously flowed directly into the computer’s memory through telephone lines to the users, making it immediately available to the operators.

  The SAGE system inspired a few thinkers, including Licklider, to see computing in an entirely new light. SAGE was an early example of what Licklider would later call the “symbiosis” between humans and machines, where the machine functions as a problemsolving partner. Implied in this symbiotic relationship was the interdependency of humans and computers working in unison as a single system. For instance, in a battle scenario, human operators without computers would be unable to calculate and analyze threats quickly enough to counter an attack. Conversely, computers working alone would be unable to make crucial decisions.

  In 1953 MIT decided to start a human factors group in the psychology section of the economics department, and Lick was put in charge. He recruited a handful of his brightest students and colleagues. Lick hired people based not on their doctoral work or class standing but on a simple test he applied: the Miller Analogies Test. (The test covers every field from geology to history and the arts. It requires both good general knowledge and an ability to apply that knowledge to relationships.) “I had a kind of a rule,” he said. “Anybody who could do 85 or better on the Miller Analogies Test, hire him, because he’s going to be very good at something.”

  In 1954 Lick’s group moved in with the social psychologists and labor-management experts in the Sloan School of Management. But the group’s ideas were far removed from management problems. As McGill, an early Licklider recruit, described it, he and his peers were far more interested in computers and computer memory devices as models for the versatility of human cognition. The first dissertation produced by the department under Lick’s guidance came from Ph.D. candidate Tom Marill, who had examined the subject of ideal auditory detection. (Like others who came into Licklider’s sphere, Marill would make his mark on the development of computer networking in years to come.) “Nothing like this had ever been seen before, at least not in a psychology department,” said McGill. The department was the first cognitive science department in history. “The work was experimentally based cognitive psychology, as it would be defined today, but at the time we had no proper language or nomenclature.”

  But eventually MIT administrators wanted something more traditional, and Lick failed in his efforts to expand his new department with permanent appointments. As a result, all his protégés, young and marketable, drifted off. “We did not have the sophistication to promote what we did, and we were unaware that there was anything unique about it. So MIT let it all slip away,” McGill said. Lick was a maverick, after all, and a little far-out, perhaps too much so for MIT.

  Yet Lick did not lament the demise of the group, for he had a self-professed short attention span. His interests changed often and dramatically over the years. He once advised a young friend never to sign on for a project that lasted more than five to seven years, so he could always move on to other things. And anything Lick got interested in, he plunged into in great depth.

  Perhaps the incident that most piqued Lick’s interest in computers and their potential as interactive instruments was an encounter he had in the 1950s with a smart, opinionated young engineer at Lincoln Labs named Wesley Clark. Clark was a young researcher working on the TX-2 machine, the state of the art in digital computation and the successor to a computer called the TX-0. Clark had built the TX-0 with Ken Olsen before Olsen left to start Digital Equipment Corporation.

  Clark’s office was in the Lincoln basement. One day, on his way back from the stockroom at the other end of the corridor, Clark decided to venture into a room that had always seemed vaguely off-limits. Most lab doors stood open, but not this one. It was always closed. Clark tried the door, which he was surprised to find unlocked, and entered the room. “I wandered in and back through a little labyrinth of baffles and barriers,” Clark recalled. “Off to one side was this very dark laboratory and I went in, and after probing around in the dark for a while I found this man sitting in front of some displays. He was doing some kind of psychometrics, and he was clearly an interesting fellow. I got interested in what he was doing and in his apparatus, and as I recall I suggested to him that he could achieve the same results by using a computer.” The man was Licklider. Clark invited Lick to come down the hall to see the TX-2 and learn some fundamentals.

  Teaching Lick actually to program the machine would have been too difficult. Programming a computer like the TX-2 was something of a black art. The TX-2, which contained 64,000 bytes of memory (as much as a simple handheld calculator today), took up a couple of rooms. What many years later became tiny microchips for the computer’s central processing unit were, in those days, huge racks of many separate plug-in units, each consisting of dozens of transistors and associated electronic parts. Still more space was taken up with large consoles covered with switches and indicator lights to help the operator or troubleshooter understand what the system was doing. All of this equipment required rack upon rack of gear, only a tiny fraction of which—the video display screen and keyboard—might be recognizable today as ordinary computer parts. “To sit at the TX-2 with Lick was to be embedded in a welter of seemingly irrelevant stuff,” Clark said. To become a TX-2 “user” would have been a daunting exercise even for someone as quick as
Licklider. For one thing, there were no teaching tools per se, no instructional aids or help menus. For another, the operating system, which would standardize programming for the machine, had yet to be written.

  One thing the TX-2 did do very well was display information on video screens. That made it one of the earliest machines for interactive graphics work. It was this feature that helped Clark demonstrate for Lick the main ideas of interactive use.

  The sessions with Clark made an indelible impression on Lick. He drifted further from psychology and toward computer science. As his interests changed, Lick’s belief in the potential for computers to transform society became something of an obsession. Succumbing to the lure of computing, he began spending hours at a time at the interactive display console. Louise believed that if he weren’t being paid for this work, he’d have paid to do it.

  The idea on which Lick’s worldview pivoted was that technological progress would save humanity. The political process was a favorite example of his. In a McLuhanesque view of the power of electronic media, Lick saw a future in which, thanks in large part to the reach of computers, most citizens would be “informed about, and interested in, and involved in, the process of government.” He imagined what he called “home computer consoles” and television sets linked together in a massive network. “The political process,” he wrote, “would essentially be a giant teleconference, and a campaign would be a months-long series of communications among candidates, propagandists, commentators, political action groups, and voters. The key is the self-motivating exhilaration that accompanies truly effective interaction with information through a good console and a good network to a good computer.”