Computing and the National Science Foundation, 1950-2016. William Aspray

who was responsible for guiding many of the research efforts funded by OSI and its successor, the Office of Science Information Services.

While OSI’s primary mission was managing and coordinating science information across federal agencies, NSF also began to support applied and basic research activities. In May 1956, NSF sponsored a meeting¹³ of representatives of the Department of Defense, National Bureau of Standards, and the Patent Office, as well as experts in linguistics, logic, information theory, operations research, computer design, and library science, to discuss fundamental research on the organization of information. On April 15–17 of the following year, Western Reserve University (WRU) hosted a Symposium on “Systems for Information Retrieval.”¹⁴

Following a period when Thomas Jones was acting head, Burton Adkinson became head of OSI.¹⁵ As Adkinson noted, “In 1957, two unrelated events made a big impact on NSF/OSI. The first was the untimely death of Alberto Thompson, who had barely started to develop a vigorous scientific information support program. Second, the launching of Sputnik surprised most Americans.”¹⁶

In 1958, the President’s Science Advisory Committee (PSAC) created the “Baker Panel.”¹⁷ Packed with luminaries and influential figures,¹⁸ this panel issued a report on “Scientific Judgments on Foreign Communications Intelligence” that called for improving the availability of U.S. scientific and technical information.¹⁹ PSAC endorsed the recommendations of the Baker Report; and the President’s Science Assistant, James Killian, Jr., urged presidential approval. A White House press release in December 1958 directed “the National Science Foundation [to] take leadership in bringing about effective coordination of various scientific information activities within the Federal Government.”²⁰

The post-Sputnik National Defense Education Act (NDEA) became law on September 2, 1958. It contained major provisions²¹ for loans to higher education students; fellowships for advanced study of mathematics and science; guidance counseling and testing to identify able students; improvement of K–12 science, mathematics, and foreign language programs; vocational programs; and research on effective uses of television and other media for educational purposes. In addition, the NDEA authorized the National Science Foundation to establish a Science Information Service: first to address indexing, abstracting, translating, and to provide other services leading to a more effective dissemination of scientific information; and next to undertake programs to develop new or improved methods for making scientific information available.²²

On December 11, 1958, NSF established the Office of Science Information Service (OSIS) with Adkinson as head. By the end of the decade, OSIS had made 146 grants totaling about $3.8 million under four major programs: Documentation Research (through which most of the research and development was funded), Foreign Science Information, Publications and Information Services, and Unpublished Research Information. Among these grants²³ were projects on linguistic transformation for information retrieval at the University of Pennsylvania and mechanical translation projects at Harvard Computation Laboratory, Georgetown University, the University of California, Massachusetts Institute of Technology, and the Cambridge Language Research Unit in England. OSIS also funded the National Bureau of Standards to establish a Research Information Center and Advisory Service on Information Processing in 1959.²⁴

In the late 1950s, it was unclear how to classify the various fields that encompass the basic sciences behind computing, computers, information, communications, and the fields that depend on them. Louis Fein, a Stanford Research Institute (SRI) consultant, was asked by Frederick Terman and Albert Bowker of Stanford University to design a computing curriculum. Fein began studying university programs “in the fields of computers, data processing, operations research, and other relatively new and apparently closely related fields.”²⁵ His goals were to identify not only computing-related organizations, curricula, research programs, and facilities, but also computing-related fields of study, and the role of the universities in these fields. As Fein noted in 1959,²⁶ “universities, as institutions, are having a hard time . . . learning how to effectively incorporate these new fields into the academic structure.” In recommending the creation of a Graduate School of Computer Sciences at Stanford, Fein defined two research-oriented departments.²⁷ “Information and Communication” encompassed instruction and research activities in information theory, switching theory, coding theory, automata theory, artificial intelligence, learning, language translation, and theory of simulation. “Systems” comprised instruction and research activities in management science, econometrics, systems theory, information classification, indexing and retrieval, model theory, self-organizing systems, and adaptive mechanisms. Today, the former might fall under a computer science (or engineering) department, while the latter might be divided among departments of information systems, information technology, and management information science. Fein saw a divide between the science of computing, communications, and information and the application and use of computing, communications, and information.

As we describe in Chapter 2, efforts to formally establish computer science as a discipline accelerated in the late 1960s and early 1970s. By the early 1960s, the fields and practitioners of information technology and information science were becoming better defined. Information technology—the more applied side— was staffed by information specialists, while information science—the research side—was staffed by information scientists. As we relate later, Altman and Brown²⁸ described the creation in the 1980s of the CISE Directorate as a move away from the library scientists and specialists supported under OSIS, to support for computer and information scientists.

Dorothy Crosland organized a series of conferences²⁹ at the Georgia Institute of Technology, for the first time making a distinction between information specialist and scientist. A specialist was someone who applied technology to the storage, indexing, and archiving of information, while a scientist was concerned with the nature of information and its representation. These conferences had a significant impact on the establishment of new information research programs at Georgia Tech, Lehigh University, and Drexel University.³⁰

The OSIS programs continued to expand. In 1967, OSIS made grants to Georgia Tech (Vladimir Slamecka) and Ohio State University (Marshall Yovits) to expand programs in information science. It also made grants to professional scientific societies to improve their literature services. The Georgia Tech center had two principal activities: mathematical models for information in the scientific disciplines and control of information for problem solving and decision making in an academic environment.³¹ By 1968, NSF awards to various professional societies to develop computerized information retrieval systems had grown to $17.7 million,