Transforming Computer Technology: Information Processing for the Pentagon, 1962-1986 (review)

The cliché of the information age is that things happen in a compressed time scale, with one year of “Internet time” the equivalent of six or seven years of “normal time.” Transforming Computer Technology, the outcome of a study led by the former director of the Charles Babbage Institute at the University of Minnesota, Arthur Norberg, confirms that feeling. The study began in the mid-1980s as an idea of the director of the Information Processing Techniques Office (IPTO) of the Advanced Research Projects Agency (ARPA) of the Defense Department. An interim report appeared in 1992, and this book in 1996. Although ten years from the beginning of research to the publication of a final report is not unusual, for this topic it seems an eternity. Around 1990 the Internet, inspired by ARPA’s work on networking, exploded onto public consciousness. Between the appearance of the interim report and this volume, the World Wide Web was conceived and became a new medium of communication. With that has come a torrent of “histories” of the Internet, ARPANET, and the Web. Not surprisingly, many of these are available only on the Web itself, its authors believing that to publish them on paper would take too long in Internet time to make the history worthwhile.

In the midst of this phenomenon appears this volume, written by trained historians who followed established methods of scholarly research. Their account of ARPA is based on primary documents made available by that agency, supplemented by research in the secondary literature and interviews with many key participants. A strength of the Charles Babbage Institute staff is its competence in conducting oral histories, especially in knowing the need to cross-check what an interviewer says. I should not have to mention that, except that many other histories of ARPA and the Web do not respect such protocols. In short, one can trust the chronology presented in this book. I should not have to mention that, either, but I must in the face of assertions that the Web has become an acceptable—even the preferred—place to do scholarly research. That may be true. But one takes a risk when searching the on-line histories for factual information.

The authors’ interpretation of those facts is another matter. It is their thesis that “IPTO program requirements had a primary influence in changing the style of computing, bringing us the computing world of today” (p. 294). They argue that the demand for batch computing with punched cards was so great that computer manufacturers had no incentive to innovate, other than to make “more of the same.” Only a strong vision from the IPTO could change that direction, and that was neither inevitable nor obvious at the time. The authors argue that IPTO’s first director, J. C. R. Licklider, had the necessary vision, and he convinced ARPA to supply the money. All of us who use computers today, especially if we use the Internet, feel the effects of that vision. The armed services, for whom ARPA was set up, have felt this effect too—the experience of the 1991 Persian Gulf War may be considered one result of this effort.

Transforming Computer Technology begins with a somewhat muddy administrative history of ARPA and IPTO, which is relevant to the thesis but seems unconnected to what follows in the book. The story really begins in chapter 2, with a chronicle of IPTO’s influence on “time-sharing”—moving computing away from the punched cards and long turnaround times that defined it in the 1950s. Subsequent chapters cover graphics, networking, and artificial intelligence (AI). A final chapter sums up the impact of these efforts.

The chapters are well written and persuasive, but uneven. Those on time-sharing and networking are the best. The idea of time-sharing occurred to researchers before IPTO entered the scene, and some form of it might have emerged anyway...