Mina Rees

The Office of Naval Research (ONR) played a vital role in the postwar support of basic research in the sciences at the nation's universities in the years immediately following World War II before the establishment of the National Science Foundation in 1950 and its expanded funding a few years later. Part of ONR's mathematics program was devoted to the support of computers and computing, and this article will report on that aspect of ONR's activities and the related activities at the National Bureau of Standards (NBS).1 The establishment of ONR at the end of World War II reflected the concern of several of the naval officers who had been associated with the work of the wartime Office of Scientific Research and Development (OSRD) and of leaders in both the legislative and executive arms of the government that the vitality and momentum of wartime research would be lost in the postwar years and the level of civilian scientific research would be disastrously diminished. This concern was formulated by Secretary of the Navy James V. Forrestal in a memorandum he sent to the president, dated February 1945, in which he said, "The problem which began to emerge during the 1944 fiscal year is how to establish channels through which scientists can [contribute to the nation's security by carrying on research] in peace as successfully as they have during the war." In response to this initiative and the support it received, ONR was established by act of Congress in 1946. In June of that year, I was invited to go to Washington to set up its program in the mathematical sciences. I had been executive assistant to Warren Weaver, the chief of the Applied Mathematics Panel (AMP) of OSRD during the war, and had frequent dealings with several of the naval officers attached to the U.S. navy's liaison office with us. It was one of these officers who suggested my name to Alan Waterman, the first chief scientist of ONR whom I had known in OSRD. Although I thought it somewhat unlikely that mathematicians would be enthusiastic about accepting support for their peacetime research from one of the military services, I decided, after talking with some of my wisest friends, to accept the appointment. During the early days of my tenure, I traveled widely around the United States to discuss with senior research mathematicians the type of program that would be desirable and appropriate, and their willingness to participate in it. The only guidelines laid down by the navy for the determination of this program were those set by Public Law 588 of 1946, which, in the words of the preamble, established an Office of Naval Research in the Department of the Navy; to plan, foster, and encourage scientific research in recognition of its paramount importance as related to the maintenance of future naval power, and the preservation of national security; to provide within the Department of the Navy a single office, which, by contract and otherwise, shall be able to obtain, coordinate, and make available to all bureaus and activities of the Department of the Navy, world-wide scientific information and the necessary services for conducting specialized and imaginative research; to establish a Naval Research Advisory Committee consisting of persons pre-eminent in the field of science and research, to consult with and advise the Chief of such Office in matters pertaining to research. The program we designed after wide consultation received the blessing of Captain Robert D. Conrad, USN, the military officer who was responsible for the entire contract research effort of ONR and was viewed by the staff as the spiritual father of ONR. In 1947 he wrote to me that I had "won the title of mathematical architect of O.N.R." The program we planned provided for the support of research in pure and applied mathematics, statistics, and computer development with its related numerical analysis to ensure the sophisticated use of electronic computers when they became available. Implementation of this broad policy was in the hands of the staff, and an able staff was assembled. The objective was to develop a program of first-class mathematical research that was of long-range interest to the navy and to identify mathematical results of potential importance to the navy wherever and whenever they appeared (see [21]). One such result that appeared very soon was George Dantzig's work in linear programming, which, as I will report later, gave rise to a substantial and important logistics program in ONR.