Dr. Adkinson and NSF/OSIS leadership

Those of us fortunate enough to receive funding from the National Science Foundation's Office of Science Information Service (NSF/OSIS) in the halcyon days of the late 1950s and early 1960s when Burton W. Adkinson, Helen Brownson, Eugene Pronko and others were there will never forget it. The financial support was not trivial, but it was not as important as the moral and intellectual support we received from them as we groped toward solutions for the information explosion facing every scientific discipline.

I say moral because we had to believe that what we were doing was in the national interest, not just for our personal growth and academic achievements-in fact, most of the research funded by OSIS in those days was being done outside academe, in professional societies, think tanks and government agencies. I say intellectual because we were guided by the vision of the OSIS staff to come together to evaluate each other's work, to work together across projects to accumulate knowledge more rapidly and to come to appreciate what different perspectives could bring to mutual problems.

I wish I had a nickel for every plane ride I took between New York City and Washington, DC, during my six years at the American Institute of Physics (1961-1966). The rides were almost weekly, and each one was important because of the people I would meet and the work I would hear about at NSF sponsored get-togethers. The trip that stands out the most in my mind after all these years is a meeting on evaluation, held at the American Chemical Society, chaired by Ed Bryant and Don King, with Cyril Cleverdon telling us about his Cranfield results. By the end of the day we were able to convince him not to use relevance as a measure but to call his measures precision and recall, and he had convinced us that we needed to do better design of our evaluation studies.

Many of the meetings went that way. Oftentimes OSIS would sponsor symposia--at Rutgers, for example, in the early 1960s-to help us all scrutinize any organization schema from all over the world: SYNTOL from France, Colon Classification from Ranganathan in India, Universal Decimal Classification from Europe, etc. How can you adopt one of these for use in the new computer environment, the NSF folks would say, if you don't critique it?

We witnessed the power of the NSF/OSIS to provide leadership and guidance as different disciplines made forays into areas of information retrieval research. The conference on Machine Translation and Information Retrieval in 1965, where the results after many years of funding were shown to be nil as far as practical information systems were concerned, meant the end of NSF funding in that area for many years. The focus of NSF/OSIS in those days was to help the professional societies in science and engineering improve the information services which they provided. Results were forthcoming in chemistry, physics, psychology and other disciplines. These developments dovetailed with mission oriented efforts at NASA, AEC, DOD and elsewhere in the government.

Without the leadership of NSF/OSIS, the efforts would have been as dispersed as later developments have been in the private, academic and government sectors-where individual projects lead to tenure, dissertations and promotions, or lead to new commercial products, new customers or new contracts, rather than new knowledge. NSF/OSIS understood that they needed to foster information exchange and communication between and among the various funded projects. That is something very worthwhile and not easily performed at meetings of professional societies, which tend to disperse rather than concentrate information about ongoing information retrieval projects. The editions of NSF/ OSIS's Research in Progress were invaluable as a way to find out who was doing what and where before anything was published or announced at some meeting. The OSIS staff seemed to understand the true meaning of information retrieval-before computers, before the Internet, before the World Wide Web and before fulltext retrieval.

Copyright American Society for Information Science Oct/Nov 1995
Provided by ProQuest Information and Learning Company. All rights Reserved