Computing Research Association: Twenty-Five Years of Service

Tonight, I would like briefly to review the history of Computing Research Association and its twenty-five years of service to the computing research community.

Let me begin by reviewing the climate in which CRA was created. The first computers were an outgrowth of the Second World War, and the first computer courses were taught in U.S. universities soon after the war ended. The major professional societies, as well as the first journals and conferences, were founded in the early 1950s. A computer industry had gelled by the late 1950s. DARPA and NSF's formal computing programs were established in the 1960s. The first doctoral programs in computer science were begun in the early 1960s, followed later in the decade by model curricula for the many new graduate programs. In the technology area, timesharing was spreading quickly, and the first work was underway on networking. One can fairly say that, by 1970, the computing professional community had been established.

However, there were still unmet needs, felt most acutely by the academics. ACM and the Computer Society were both active organizations, but they had many interests to serve. The majority of their members were programmers and others working in non-research positions in industry. They had some programs valued highly by the academic computer scientists, but the department chairs felt a need for more attention devoted to curricular issues, doctoral production, academic salaries, and research directions.

In 1970 Earl Schweppe, a computer science professor at the University of Kansas and one of the developers of the first ACM model curriculum in computer science, held a conference at his university on research issues in computing. This conference attracted many of the key computing researchers in the United States and demonstrated the value of this kind of conference. In the November 1971 issue of Communications of the ACM, Penn State computer science professor Preston Hammer called for a new winter meeting that would serve both for research communication and job placement. The following year, a group of chairs from Ph.D.-granting computer science departments, mostly located at midwestern universities, met to discuss issues concerning doctoral production and research issues. This was the first of a series of such meetings, most of them held in Columbus, Ohio and organized by Marshall Yovits, the chair at Ohio State.

The discussions of these department chairs led to the establishment of the first Computer Science Conference. This effort received early encouragement from ACM, but the chairs could not originally come to terms with ACM on running the conference, so they organized the first two meetings on their own, with financial support from the National Science Foundation. The first of these conferences was held in March 1972 in Columbus. For several years the conference was run jointly by the department chairs and ACM, but then it was taken over entirely by ACM as its major research conference. These conferences were held annually into the 1990s, although they have been abandoned now. One key element was an employment registry, which was run by Orrin Taulbee, the chair from the University of Pittsburgh. The department chairs held their own meetings and organized panel discussions or workshops on issues concerning department chairs at these annual conferences.

Because the department chairs had interests that extended beyond a research conference, they formalized their organization as the Computer Science Board (CSB). Their charter called for them to organize the chairs of computer science departments, promote communication among them, and provide a forum for discussion of ideas. They set a ratio of representation on the board of three-quarters academic and one-quarter industrial representation, which is still followed today.

The intention from the beginning was to make this a North American rather than solely a U.S. venture. There has never been much participation from Mexico, but Canadian computer scientists have been active from the outset. Arthur Wouk, from the University of Alberta, was a member of the original board. For many years, John Brzozwski from the University of Waterloo was active in the organization, serving a few years as an officer. John Tartar from the University of Alberta was chairman of the board and a key figure throughout most of the 1980s. Maria Klawe from the University of British Columbia and Ken Sevcik from the University of Toronto have been active board members in the 1990s.

The number of departments was growing rapidly in the 1960s, and the CSB decided as one of its first projects to keep track of departments that granted doctorates in computer science. Later this project was expanded to track doctoral programs in computer engineering as well. The list of departments was named after George Forsythe, the founder of the computer science program at Stanford University. The keeping of this list continues today.

CSB also decided that it would be helpful if the computer science department chairs could periodically come together and discuss their concerns. Computer science was a new and growing intellectual discipline with many growing pains. Tony Hearn from the University of Utah identified a sleepy resort in the Wasatch Mountains that was a lovely place to hold conversations while you hiked the mountain trails. The first of these conferences, organized by Marshall Yovits , was held at this resort in Snowbird, Utah in 1974; and the conference has been repeated alternate summers since then. During the 1980s, most of the conferences resulted in a white paper on the state of computing.

With the rapid expansion of academic computing in the 1960s, the community wanted to be able to track its progress through better data collection. For example, Tom Keenan from the University of Rochester and John Hamblen from the Southern Regional Education Board collected information about computers on campus. Taulbee began to collect information about doctoral production in 1970, and his data project was turned over to CSB when it was formed in 1972. In the 1970s and 1980s, results of the annual Taulbee survey were published in Communications of the ACM, and sometimes in the IEEE's Computer magazine as well. In the early 1990s, our present practice of publishing the results in Computing Research News began.

During the 1970s and 1980s, CSB addressed a number of issues of concern to computer science departments. These included the mismatch in the growth in the profession and the federal funding for research, the charge structure for computer time in academic departments, the shortage of good referees for the professional journals, inadequate production of doctorates, faculty salaries, recruitment and retention of graduate students and faculty, curricular issues, computing infrastructures, accreditation efforts, and immigration issues.

Throughout the 1980s, there was talk that the computer science community did not have adequate representation in Washington. The biologists, chemists, and especially the physicists had this kind of representation and used it effectively in lobbying for additional federal funding. This concern of the computer scientists was shared by professionals in Washington-especially at NSF and at the National Academy of Sciences-who were interested in the state of computing and its national importance. Serious discussion of this issue was held at the 1986 conference at Snowbird.

As a result of this conference, several changes were made to the Computer Science Board in 1986. It began to position itself to act as the community-wide representative for computing research. It incorporated as a not-for-profit educational organization. It changed its name to the Computing Research Board (CRB) so as to include computer engineering as well as computer science. It changed its method of election, having board members elected by the chairs of Ph.D.-granting departments instead of by the board itself, so as to be more responsible to the computing research community.

At the 1988 conference at Snowbird, the problem of national representation was again seriously considered. There was still no organization in Washington that represented the computing research community on an ongoing basis. CRB recommended to the department chairs that the departments be assessed annual dues so that a staff could be hired to provide a permanent presence in Washington. The departments were overwhelmingly supportive and agreed to assess themselves dues sufficiently great that the work could be done well. Start-up funding was raised from ACM and a number of companies (AT&T, DEC, IBM, and others). Apple provided computers for the office. The Computer Society provided the original office space and furnishings. The name of the organization was again changed-to Computing Research Association (CRA).

Terry Walker from the board served as interim director until the board was able to hire Frederick Weingarten to be both executive director and director of government affairs. Weingarten had many years of experience at NSF and the Office of Technology Assessment, which he brought to bear for CRA. During his tenure as executive director, he made CRA an effective presence in the Washington policy community and developed the programs and finances of the association. After five years, CRA had grown to a point that it was time to separate the executive director and government affairs director positions. Weingarten chose to remain in the government affairs position as a part-time employee, and William Aspray, a computer historian and research center administrator, was hired in 1996 as the new executive director. Other staff members were hired to work on publications, web pages, programs, and administration.

The government affairs program blossomed. CRA is today regularly consulted on issues concerning computing research policy-by staff and committees in both the Executive and Legislative branches. CRA was involved, for example, in supporting the High Performance Computing Act and the Gore initiative to expand the research and networking components of the HPC Act. CRA also organized the National Information Infrastructure Workshop in 1993, where 400 industrial and academic researchers set a research agenda largely adopted by the Clinton administration, and the Next Generation Internet workshop in 1997 to set a research agenda for national networking initiatives.

In the early 1990s CRA developed a collection of projects to maintain and increase the number and status of women participating in computer science research and education at all levels. These programs included the Grace Hopper Conference for women in computing, career workshops, summer research experiences for undergraduates, databases of women researchers, a series of articles in Computing Research News, career booklets, and various mentoring activities. These programs have been highly effective and, together with the government affairs activities, have been the major programmatic additions of the 1990s.

A planning effort in 1995 led to a new mission statement, with four major mission areas: community building (e.g., the conference at Snowbird), human resources (e.g., women in computing activities), information (e.g., Taulbee survey), and policy (e.g., NGI workshop). The structure of the organization and the dues structure have remained mostly unchanged over the past decade. There has been a significant increase in membership, especially of research labs, research centers, and professional societies. There have also been a number of new programs: Executive Fellowship, Industrial Salary Survey, Academic Profiles Survey, IT Worker Shortage Study, Leadership Summit for computer professional societies, and a minorities in computing program. We look forward to building on these programs to continue to serve the computing research community in coming years.