Where Wizards Stay Up Late: The Origins of the Internet
By Katie Hafner and Matthew Lyon
(Simon & Schuster; 320 pages; $24)
Were it not for a coin tossed in the fall of 1962, the Internet might not exist. The winner of the toss, MIT computer scientist Joseph Carl Robnett Licklider (or "Lick," as he insisted on being called), got the top job at an obscure government agency known as the Command and Control Research Division, a division of the Pentagon's Advanced Research Projects Agency (ARPA), which was created after Sputnik to close the purported gap between Soviet and American science.
When Licklider arrived, his agency specialized in the sort of esoteric Department of Defense research that could only be called "asinine" (his word, in fact). Air Force intelligence, for instance, wanted to use the agency's huge mainframe computers to detect patterns of behavior among high-level Soviet officials. The idea was to feed the computers bits of information like, "The Soviet Air Force chief drank two martinis yesterday," or, "Kruschev isn't reading Pravda on Mondays," from which the computer was to deduce that the Soviets were building a new missile, or that some inner-circle coup was imminent, or whatever.
This is a test of Word's abilities.
Licklider got rid of the fanciful war games, substituting something that must have sounded equally bizarre--the creation of an "Intergalactic Network." In two short years on the job (he returned to MIT in 1964), Licklider poured enough money into America's universities to influence an entire generation of computer scientists, of whom some would go on to invent video games, the mouse, the metaphor of "windows" and "icons" and, of course, the Internet itself. Licklider's prescience stemmed from his pioneering research on "time-sharing"--the then-unthinkable notion that more than one person could share a computer at the same time, through terminals. "Time-sharing" had the added benefit of letting people get their hands on keyboards, which few of them had ever done. Computer science up to that point had been dominated by "batch-processing," a system believed to be maximally efficient, wherein programmers wrote code at their desks, copied it onto stacks of cards with holes punched into them, and handed them over to technicians who had a priestlike say over who got to use the computer, and when.
Considering his achievements, J.C.R. Licklider may be the most influential little-known person in the history of computer science. Katie Hafner and Matthew Lyon, authors of Where Wizards StayUp Late: The Origins of the Internet, dedicate their book to him, although that, unfortunately, is the grandest of their corrective gestures. The short 15 pages they devote to his life do little more than summarize his résumé.
This reflects a larger flaw in the book. Hafner, a contributing editor at Newsweek and co-author with John Markoff of Cyberpunk: Outlaws and Hackers on the Computer Frontier, and Lyon, her husband and an assistant to the president of the University of Texas, have written an epic technological history. Their story of how the Internet was discovered is a blow-by-blow saga of how a single engineering problem, in this case the construction of the ARPANET (precursor to the Internet), was solved. But the larger intellectual revolution undergirding the Internet--the redefinition of the relationship between people and computers, which turned impersonal calculating machines into intimate tools for everyday thought--earns only a few mentions in passing. There is merit to Hafner and Lyon's narrowly focused approach. They've conducted extensive interviews with the scientists involved, culled material from archives and private collections, and revised the historical record where necessary.
For instance, they debunk the long-standing myth that the Internet was created as a test to see if computer networks could be designed to survive a nuclear war. Instead, as Lyon and Hafner show, the Net was built because the federal government needed a way for all the incompatible computers being used for government research to communicate with each other. In a sense, the Internet was the ultimate time-sharing project: a nationwide network capable of connecting any sort of computer with another computer.
T he logistics were daunting. How on earth were scientists to create connections between dozens, let alone hundreds, of computers without running up a fortune in telephone bills? The answer, which occupies three chapters of the book, is "packet switching," and it remains the core of today's Internet. Instead of opening a direct connection with the destination computer, messages are sent willy-nilly, chopped up into little packets that are reassembled when they get to their destination.
Hafner and Lyon tell the story of how two scientists, Paul Baran in America and Donald Davies in England, arrived independently at the notion of packet switching. They trace the web of conversations that brought their ideas to the office of Bob Taylor, Licklider's successor. Taylor didn't have time to manage the details of the ARPANET project, and hired one particularly brilliant graduate student named Larry Roberts to do it for him. Roberts assembled his fellow graduate students from universities around the country, then gave them the right to do more or less whatever they wanted. Perhaps the greatest mystery in the history of the Internet is one about which Hafner and Lyon never bother to speculate--how and why the Pentagon, at the height of the backlash against the Vietnam War, was persuaded to grant a bunch of longhaired, Tolkien-reading grad students huge sums of money, without insisting on oversight. Whatever the reason, it worked: The crucial design decisions were made at a series of informal grad-student conferences, and by September 1969, a working connection had been opened between a computer at UCLA and one at Stanford.
The second half of Where Wizards Stay Up Late describes the aftermath of the ARPANET project. We read about how the students developed interactive games, created the first online discussion groups, fought over what technical standards to adopt, and ran headfirst into issues still relevant today--privacy online, the appropriate use of computers, ownership of software, how to govern the Internet. Once again, though, the authors don't interpret the picture they're painting in such studious detail. The Internet, barely in the government's hands in the first place, was hijacked by the students, who seemed to be more interested in playing Dungeons and Dragons than in using the network solely for research.
How does the architecture of the Internet reflect the strange counterculture these students conjured up out of Pentagon funds? Readers who want a more complete picture should read two other books: Steven Levy's Hackers: Heroes of the Computer Revolution (1984), which gives a still-fresh account of MIT's Project MAC (not coincidentally funded by Licklider), an enterprise dominated by the "hacker ethic"--the compulsive urge to explore and improve on things. (In that sense, the Internet was the ultimate hack.) And Stewart Brand's II Cybernetic Frontiers (1974), which recounts how this generation of "computer kids" designed computer games based on science fiction and used the Internet to fashion a universe of their own. Building the Internet was not simply about building a better mousetrap; it was, for its creators, also about building a better and smarter, and wildly utopian, world. That story remains to be told.