CHARLES BABBAGE INSTITUTE
|Paul Armer (1924–2016)|
We were saddened to learn of Paul Armer’s passing early in 2016. He had a prominent and varied career in computing, in the nation’s social and political discussions about computing, and not least in helping launch the Charles Babbage Institute, for which he served as executive secretary during the crucial years when CBI was moving from California to Minnesota.
Paul’s first career in computing was with the newly formed RAND Corporation, which he joined after wartime service with the U.S. Army Air Corps, a brief stint at United Airlines, and then completing his degree in meteorology at UCLA. He published a very early RAND Memorandum in 1947 with the ungainly title “The Location of the Maximum of a Function of Two Independent Variables When the Dependent and Independent Variables are Measured Without Error.” RAND watchers will recall its founding as a spin-off from Douglas Aircraft in May 1948.
In 1952 Armer became head of RAND’s numerical analysis department. During these years at RAND he worked on some of the earliest IBM computers, such as the Card Programmed Calculator, to address aircraft design problems as well as the incidence of mental illness in the United States. He also guided RAND’s largest computer project, the JOHNNIAC based on the design of John von Neumann’s Institute for Advanced Study computer. “We decided to embark on the construction of JOHNNIAC. In early memos, I’m pretty sure Willis talked about building it for $150,000. By the time it computed its first prime number, I think we conservatively had a million dollars invested.” (Armer 1973: 16) JOHNNIAC, operational from 1953 to 1966, was at the heart of RAND’s advanced computational research in programming languages (JOSS) and air-defense systems; its early work on SAGE programming led to the spin-off of the System Development Corporation. Work with the IBM 701 and 704 computers led to Armer’s leadership in the seminal user groups GUIDE and SHARE (for which we have archival records that Paul himself arranged the donation).
Armer served as RAND department chair until 1962. As he described it, he persuaded RAND management “to let me switch hats with Willis Ware . . . and I would spend that [next] year in a semi-sabbatical way getting myself refurbished.” (Armer 1973: 5) The unit subsequently evolved into RAND’s computer science department under the leadership of Willis Ware. Armer’s activities moved sideways into engagement with social and political issues in computing, including debates on computing and automation. His RAND publications make clear his contributions to national discussions on technological change, employment, and automation; computing in the Soviet Union; and the “computer utility” movement. His 1968 testimony to the US Senate, “Privacy Aspects of the Cashless and Checkless Society,” raised prescient questions about e-banking long before the actual arrival of e-banking.
RAND during the 1960s was quite a place, experiencing a fundamental transition from its Air-Force dominated origins to a wider framing with high-visibility projects in social policy and urban affairs. Partly by accident, Armer helped expand the national conversation about artificial intelligence by hiring MIT philosopher Hubert Dreyfus for a summer in Santa Monica. “It struck me . . . that I ought to mix a philosopher in with these [artificial intelligence] guys, that there were indeed philosophical questions here. I thought maybe the project would gain from having a philosopher around for a while,” Armer told Pamela McCorduck for her book Machines Who Think (second edition 2004: chapter 9).
A RAND colleague recommended his brother, who happened to be Hubert Dreyfus. Over the summer Dreyfus drafted, then published as a RAND memo the polemical Alchemy and Artificial Intelligence (1965). Armer later described it as “lousy philosophy” but decided to issue it anyway, and, lo and behold, “it became the best seller of any such paper that RAND put out.” Dreyfus, buoyed by the prestige, went on to be one of the leading philosophically grounded critics of AI. His classic What Computers Can’t Do appeared originally in 1972 and was revised and reissued two decades later.
Also at RAND in the 1960s were two original copies of a top-secret history of the Vietnam war, better known today as the Pentagon Papers. Armer and his wife Joan, according to their daughter, were sympathetic to the anti-war movement and friendly with Daniel Ellsberg, who mailed an additional photocopy he had made to the New York Times which began publishing excerpts of the controversial documents in June 1971. The event marked a turning point in the mobilization against the Vietnam War. The full and complete 7,000 pages of the Pentagon Papers were finally released in 2011.
After his years at RAND Armer moved to Stanford in 1968, where he directed the university’s computation center for two years and then lectured in the Computer Science Department throughout the 1970s. Even in the early 1960s Armer’s interest in the social effects of computing and automation led him to do course work in economics at UCLA. In the mid-1970s he conceived a book project that he worked on at the Program on Technology and Society at Harvard University and then also at the Center for Advanced Study in the Behavioral Sciences at Stanford. A stream of essays, reports, and Congressional testimony resulted, although (so far as I’m aware) not a book per se.
When Erwin Tomash organized the International Charles Babbage Society in California in 1977-78, Paul Armer was one of the insiders in the impressive brain trust that Tomash assembled. Armer accepted the post as Executive Secretary—effectively, acting director—and opened an office at 701 Welch Road in Palo Alto (two blocks away was Sand Hill Road, the spine along which venture capital was coalescing). Armer was one of the members of the truly distinguished Advisory Committee that met April 28, 1978, at Rockefeller University under the auspices of its president Joshua Lederberg. The society was renamed “the Charles Babbage Institute” in May 1978, and with determination Armer and Tomash moved to line up support from AFIPS. We have a lovely photograph of Tomash, Armer, and Al Hoagland in June 1979 signing the papers for this historic agreement. It happened that Armer had been president of AFIPS in 1968-69 while Hoagland was the current sitting president.
Even beyond the AFIPS agreement Armer took on numerous organizational tasks for CBI, each of them momentous. Already in September 1978, Armer with a colleague did the very first CBI oral history with Gerhard Dirks, the German inventor of a notable disk drive similar to IBM’s RAMAC. The interview is explicitly mentioned in the inaugural issue of the CBI Newsletter from May 1979. Even though this interview did not see the light of day, it is somehow fitting that Paul Armer’s own oral history done in June 1980 is officially “CBI OH #1.” Armer was also at hand when the CBI Reprint series was taking form, which emerged as a major activity across the 1980s. Possibly the most important activity that he guided was the national search for a permanent university home for CBI, which resulted in CBI moving to the University of Minnesota in 1980, an effort that I have described in an earlier CBI Newsletter article.
One contribution that Armer was particularly proud of was his ability and agency in networking. He held numerous leadership positions in ACM, SHARE, and AFIPS, and sought to extend their collaboration with DPMA and other professional societies. He was a founder of Institute for the Future in Menlo Park and an insider with the trade journal Datamation. He liked, as he put it, “to sometimes go see good papers and then work like hell to get that paper published.” At RAND he was close to the Los Angeles office of Datamation, “so every editor from the beginning has always been someone I’ve had access to.” (Armer 1973: 68) He gave an instance of hearing a paper by Lance Hoffman and William Miller and shepherding that into print (see interview with Hoffman at <http://hdl.handle.net/11299/168279>). It happens that CBI’s shelf copies of Datamation, which we regularly consult for its wealth of information, is Armer’s personal copy (with his name embossed on the spine). We salute his many contributions to computing and the history of computing.
Thomas J. Misa
Paul Armer interview (1973). Computer Oral History Collection (#5) Smithsonian Institution Press Washington, DC, USA 1999. Available at <http://amhistory.si.edu/archives/AC0196_armer730417.pdf>.
Paul Armer interview (1980). Charles Babbage Institute OH 1.
Paul Armer obituary: New York Times (26 January 2016). Available at <http://www.legacy.com/obituaries/nytimes/obituary.aspx?pid=177458483>.
Paul Armer Collection: National Museum of American History. Finding guide available at <http://amhistory.si.edu/archives/AC0323.pdf>.