Computer Science and the Role of Government in Creating the Internet: ARPA/IPTO (1962-1986) Creating the Needed Interface by Ronda Hauben rh120@columbia.edu Section II Basic Research for the National Defense and the U.S. Department of Defense: A Paradox? Preface I put down...two basic principles for successful Government participation in scientific research. First, the research organization must have direct access to Congress for its funds; second, the work of the research organization must not be subject to control or direction from any operating organization whose responsibilities are not exclusively those of research." Vannevar Bush Testimony in Congress, 1945 "For me as a scientist the military has been the most comfortable group to deal with, and at present I have found it to be the one organization in this imperfect world that has the funds and spirit to advance science rapidly and successfully." Theodore von Karman "(T)he quality of life on earth tomorrow will be determined in large part by the measure of the scientific research undertaken today. There is thus a significant public responsibility to sponsor research in the various scientific disciplines and to keep the way clear to follow up on new discoveries. Determining the emphasis, however, is a most delicate responsibility...." "It must be clearly understood that most of this money purchased research of the highest quality. However, not nearly so clear is the rationale that dictated that the Department of Defense should be the principal sponsoring agency for much of this vital research." Senator Mike Mansfield The distinction between applied and pure research is not a hard and fast one...But it is important to emphasize that there is a perverse law governing research: Under the pressure for immediate results, and unless deliberate policies are set up to guard against this, "applied research invariably drives out pure." The moral is clear: It is pure research which deserves and requires special protection and especially assured support. Vannevar Bush Science: the Endless Frontier July 1945 I-A Successful Basic Research Program is Challenged by the US Congress. The question emerges of why and what would be the effect? In fiscal 1970, the Senate Appropriations Committee raised the question of whether the U.S. Department of Defense was within its mission obligations to be funding the forefronts research and development work that it was supporting. This research included the development of a new scientific field, the field of computer science, and more particularly, the field of computer communications. This question raised by Congress, whatever the cause, had an important effect on the development of this new science and of its progeny which includes the Internet among its spectacular achievements. In this and upcoming sections of this draft paper I want to explore several questions. The first and perhaps most interesting is: Why would the U.S. Department of Defense be the place where this new field of scientific research developed? I also want to look at this field of research, the field of computer communications research and explore how it developed. And perhaps most importantly, I want to look at the effect that the Congressional concern had on the research and the researchers. These experiences raise the question why the US government, which has a constitutional obligation to provide for the welfare of its citizens and for the national defense, would act in this way. Are these actions contrary to these obligations? A second significant result of this challenge by Congress would ultimately be that the Information Processing Techniques Office (IPTO) which pioneered basic research in computer science and in the new field of computer communications at ARPA in the DoD would be ended in 1986. With the end of IPTO a significant loss was sustained by people in the US and abroad who have gained much from the work of this office. However, to understand the nature of the problem that led to the end of IPTO in 1986, it is important to look back and try to understand why basic research in science and then in computer science was able to develop and thrive for an important period of time within the U.S. Department of Defense. II. Basic Research and the National Defense The actions of the US Congress raise some important questions that require serious consideration. Among these questions is the issue of: What is the nature of basic research and how did support for certain kinds of basic research become a responsibility of the U.S. Department of Defense? To answer this question, it will be helpful to review how it is that civilian conducted basic scientific research that could be considered crucial to the national security or national defense would fall within the responsibility of the U.S. Department of Defense. Before the WWII was over, there were a several different proposals for the form that any postwar research institution should take to continue civilian scientific involvement in national defense. The important scientific and technical achievements of the Office of Scientific Research and Development (OSRD) established during World War II had demonstrated the value of civilian scientific research for the public in general and for the US Department of Defense, in particular. There were discussions and proposals for how to continue the collaboration between civilian scientists and the U.S. government. Army and Navy officers involved in these discussions were not satisfied with the proposals that emerged. Nevertheless, at a conference held in April 1944, they did agree on the principle that in the "postwar era power to conduct long-range defense research should be vested in a civilian-dominated agency like the OSRD." (4) 1. The Research Board for National Security Subsequently, a committee was formed to explore this recommendation. It included Charles E. Wilson, president of General Electric, Frank B. Jewett, the president of the National Research Council (NRC) and the head of Bell Labs, Jerome C. Hunsaker who was chairman of the National Aeronautics Council of American (NACA), Merle A. Tuve, a physicist at the Carnegie Institute in Washington, and eight high ranking Army and Navy officers. Daniel Kevles in an article about this period, describes how Tuve "warned the Wilson Committee that the armed services could not hope to enlist civilian scientists in peacetime to fill a subordinate role in military research."(5) Tuve recommended that if civilian scientists were given "equal and autonomous status" with the military, they would contribute to the national defense. His proposed mechanism for doing this was to create a joint military and civilian research agency, the Research Board for National Security (RBNS). The role of the RBNS would be to explore the need for more fundamental research needed for national defense. His proposal was for an independent federal agency that would be established by Congress. Tuve's proposal met with opposition within the Wilson Committee itself. Jewett preferred that there be a private sector approach to the problem of civilian participation in national defense research. He proposed that the RBNS be established under the National Research Council (NRC) of the National Academy of Science (NAS), which was a private, nongovernmental organization (6), though created by the U.S. government. A Wilson Committee report of September 14, 1944 approved Jewett's proposal for the RBNS to be created under the auspices of the NRC as a temporary measure, and for a Congressionally authorized independent entity to be created as a more long term solution. (7) The Army representatives who were to consider this proposal, Colonels William Borden and Ralph M. Osborne were worried that if Congress got involved in the creation of the RBNS, it might begin to intrude into the military's responsibility and also absorb some of the funds that otherwise would go to the military. Therefore, they favored creating the new agency under the NRC and maintaining it as a dependent of the Services as it would be dependent on them for funding. In hearings before the House Select Committee on Post War Military Policy in fall of 1944, they gave their support to the NRC plan. (8) After several other events which seemed to favor the creation of the RBNS in the NRC, staff in the Bureau of the Budget (BoB) became concerned that the creation of the RBNS, outside of the government, would be a violation of public policy. (9) Harold D. Smith, Director of the Bureau of the Budget, advised the President against the creation of such an entity until there could be adequate study done about it. Also Smith directed that the Army and Navy not transfer any funds to the RBNS until there was an adequate chance to study the postwar research and development plans. Kevles describes the concern driving these actions about this "tendency to want authority over important governmental functions vested outside the government" by those in the Bureau of the Budget.(10) The concern was that approving of such an entity would be "approving the National Academy, including its RBNS, as the focus of the government's postwar research program."(11) In pursuing his concern in this area, Smith asked President Truman to sign letters indicating that control of military research and development remain inside government. The principle he maintained had to govern is that "governmental operations remain under governmental control."(12) In early 1945, a Congressional Committee, the Woodrum Committee raised the problem that the US code did not allow agencies established by executive order to spend government money without Congressional authorization. A bill was drafted to have Congress approve money directly and the bill was reported on favorably. However, continued disagreement about the placement of a research agency within a nongovernmental entity, and the illegality of transferring funds, or appointing government officials to an agency which hadn't been authorized by Congress, led to the end of the proposal for the RBNS on February 28, 1946. (13) With the end of the proposal for the RBNS, the Army and Navy were able to obtain funds and the authority to directly contract with universities for fundamental research. And in 1946, the Office of Naval Research (ONR) was established by Congress for the Navy to support research in lieu of the RBNS. In his article describing these developments, Kevles comments about how this solution was only to be an interim solution but continued for a longer period of time as the politics over how to support peacetime basic research in science delayed the creation of the National Science Foundation (NSF). "By 1950, when the National Science Foundation was finally established," he writes, "it contained no division for defense research. Neither the army nor the navy was prepared to relinquish its new direct relationship with the civilian scientific community, nor was the civilian scientific community eager for any change in the prevailing institutional pattern."(14) Among the advantages of having direct contact with civilian scientists that Kevles notes was that the military recognized the benefit of "the prospect of civilian scientists testifying for research and development appropriations which the armed services might otherwise be unable to obtain for themselves." (15) How then did the military treat this responsibility for basic research on issues of national security or the national defense? 2. The Air Force and the AFOSR In October, 1951, a little while after the National Science Foundation was created, the Air Force began to create a research organization in the Air Force that would later become known as the Air Force Office of Scientific Research (AFOSR). The Air force established a small staff in the headquarters of the Air Research and Development Command. Nick A. Kumons, of the Historical Division of the Office of Information of the Office of Aerospace Research, writes that "The event was of no small significance. It marked the formal recognition by the Air Force that a viable military technology had roots sunk deep in fundamental science. The Air Force was becoming increasingly aware that an intimate association, a kind of benign symbiosis, existed between science and technology. Neither science nor technology could exist without the other, and the Air Force, perforce had to live with both." (16) Kumons describes how the Air Force had also realized the need to mobilize science to be able to serve military technological development, and that this task couldn't be left to others. Kumons traces the roots of this understanding back to the fall of 1944, when General H. H. Arnold decided that the essential aspect of Air Power was "pre-eminence in research." He, therefore, reasoned that it would be necessary to put the Air Forces' R&D programs on "a sound and continuing basis."(17) General Arnold gathered a set of scientists and constituted the group, the Army Air Forces Scientific Advisory Group. Theodore von Karman, a Hungarian born aerodynamicist was named chairman. The group was asked to ignore the ongoing war and instead to consider the most recent scientific and technological advances to plan for a program for the next 20 years. (18) When Bush's report "Science, the Endless Frontier" appeared in July 1945, the response of von Karman of the Army Air Forces (AAF) was hostile. "A national program in basic research," von Karmin wrote, "was a 'necessary adjunct' to the maintenance of a strong military posture," since "every scientific development eventually finds its way into the field of military applications." (19) Therefore, he maintained that it was essential that the US government sponsor a program in basic research. And he held that the AAF should not delegate "its responsibility to pursue scientific knowledge to any other federal agency." The mechanism he proposed for such government support for basic research was a broad based program with various agencies, including the AAF able to pursue the research. He believed that the AAF should be able to expand its relations, both "spiritual and conceptual" with the scientific community. "No one should act as the only source of information between science and the AAF."(20) Kumons describes how the OSRD was ended, but that nothing took its place. Meanwhile legislation establishing the Office of Naval Research (ONR), was being passed in Congress. This gave the Navy an institutional form in which to pursue its own research. In July 1947, the National Security Act created the Department of Defense. In the upper eschelons of the Department of Defense, the Research and Development Board (RDB) was created with the authority to pursue questions of military Research and Development. Vannevar Bush was at its head, but it had no money or facilities and limited power. The Air Force, therefore, wanted to have an active basic research program like the ONR. In 1948, the Air Force Directorate of Research and Development of the USAF determined that the Air Force must "undertake such basic research not now being carried on by other [federal] agencies. (21) In February 1949, after several reorganizations, the Air Materiel Command, which was earlier charged with Air Force R & D, was renamed the Office of Air Research and moved from under the Engineering Division to be an organization in parallel with it. The Air Force had decided to engage in basic research, but with no guaranteed budget or facilities, the office was not able to enforce the decision. In the meantime, there was a concern within the Department of Defense about having each Service establish its own research organization. In February 1948, there had been a proposal from the Research and Development Board headed by Dr. Karl T. Compton that a new civilian research agency be set up similar to the OSRD and it would gain control of all research in the Department of Defense. Von Karman was active in detouring this proposal. (22) The Air Materiel Command, which at this time was in charge of research and development, was not able to function because it was part of the organization also charged with procurement, maintenance and supply. Without a separate budget, R&D was at a loss to be able to argue its case before more powerful interests in the Air Materiel Command. Therefore the program only got what it was given. By 1949 Senator Stuart Symington was asked to separate the R&D budget from procurement, maintenance and supply.(23) A civilian committee headed by Dr. Louis Ridenour, the Dean of the Graduate School of the University of Illinois, concluded its deliberations in September 1949. It stressed the importance of basic research for the Air Force. The report argued that the research should be in broad areas of interest to the Air Force, "it would be in broad general fields and not directed toward definite goals or applications. The research contract itself," the report maintained, should not specify what was to be investigated, "except in terms proposed by the investigator." (24) And the contracts were to be awarded "less with regard to the description of the project than with regard to the ability and promise of the principal investigator." The research should be done at universities, "the great centers of fundamental research." In January 1950, the Air Research and Development Command was established and devoted to the problems of research and development. Kumons describes the difficult set of circumstances in the Air Force faced by those trying to create an R&D program. In the process of their effort, there was a recognition that a basic research program was different from other things that the Air Force had experience with. Kumons explains: (A) basic research program was not readily amenable to very deliberate planning. One could not sit down and logically construct a program and then go out and requisition the research he wanted done. In point of fact, this was the way the Air Force had been accustomed to doing things. Requirements were established; if it could be found, as each requirement was broken into bits and pieces, that existing materials and techniques were inadequate to the task at hand, the laboratories were simply instructed to come up with the necessary materials and techniques. This approach had its place, but it also had its limitations. The trouble was that when it yielded a solution it was only to the assigned problem. A well-conceived basic research program...would lend itself to a broad range of applications. But to do so it could not be slavishly subservient to requirements. And herein lay another difficulty with the traditional approach. Basic research was an attempt to grapple with the unknown. What its results might be was virtually unpredictable. How, then, could requirements be established for unknown elements? If anything, the reverse was more logical: the Air Force's future requirements might very well be suggested by discoveries in fundamental science.(25) Another difference between basic and applied research and technology had to do with the inspiration for it. It wasn't utilitarian in purpose. "Scientists sought new knowledge not so much as a result of external pressure but from an inner need. Since the quest for knowledge sprang from within, it followed that the scientist, the man doing the research, was the best man to decide what research should be done." (26) Unsolicited proposals would be accepted, but they would be funded on the basis of the promise of the scientist not the project. The only requirement was that the area the scientist was working on had to be one that was of interest to the Air Force. An administrator in the program explained: I don't have to manage the scientist. As soon as I determine that he is working in an area of interest to the Air Force...I need only assure myself that he is one of the best men in this field. Then it is simply a manner of letting him work on what he wants to work on--what he'd work on anyway, if he had the money. I don't have to go out to make him work, don't have to check on him every month to be sure he's doing what he doesn't want to. I need only visit him once in a while to get his input. (26) How to obtain input was to be according to the way that Faraday advocated: "...there are three stages of research: to begin it, to complete it, and to publish it." And to publish it meant in the learned journals. Too often required reports would seem slipslod, but when someone wrote for a professional journal, the person would be under the pressure of having the eyes of one's scientific peers focused on one's work. Also recognizing that the scientific community is an international community, there was an effort to support research in NATO countries. (27) Implementing such a program was, however, fraught with difficulty. Concern in the Bureau of the Budget for overlapping research programs, led in 1953 to pronouncing the National Science Foundation the primary agency for the support of basic research, and limiting other agencies to "such additional research as may be directly related to the solution of problems for which they have statutory responsibility." (28) This generated opposition among the other agencies that carried out basic research in the federal government. On 17 March 1954, Executive Order 10521 was issued, restricting agencies other than the NSF to "basic research in areas that are closely related to their missions."(29) Kumons documents the subsequent problems in the Air Force Research Program which included being lowered in the administrative structure. (30) Also the budget for Research and Development could no longer be defended on the basis of its merits. "To do so...would make it appear as if AFOSR were competing with the National Science Foundation for funds-- something which the Bureau of the Budget would not countenance."(31) Kumons shows the change in semantic categories where basic research and applied research were dropped from the terminology in budget defenses and instead exploratory research categories were adopted. He writes, "Exploratory research, which now became the sole business of AFOSR, was little more than basic research under a new guise (although it would shortly take on a unique meaning).....(A)ny and all line items that sacked of ivy and ivory towers were blotted out....In briefings, conferences, and other points of personal contact, they began speaking the language of applied research. The effort worked. AFOSR talked of applications, and the Bureau of the Budget loosened the purse strings. (`We sold them the sizzle,' cracked one AFOSR administrator, `not the steak.')...(They) had built up a semantic bridge between what was applied in AFOSR's mission and what AFOSR was actually doing."(32) Kumons describes how this change in the semantics not only helped to confuse those dealing with the budget, it also created a barrier with regard to others. "For one thing, it permanently introduced an element of confusion in AFOSR's mission, and, indeed, in the mission of other...centers. It was an exceptional individual who could now pick his way through this formidable semantic maze. Thus, what was to serve as an effective smoke screen between AFOSR and the Bureau of the Budget also served to confuse AFOSR's friends. And, as a corollary to this, it scarcely served to educate anyone to the overriding benefits of basic research to military technology."(33) Kumons shows that what was at first a semantic change in the nature of the program, soon came to be an actual change.(34) As the defense of Air Force research and development programs could not be on the basis of the requirements of basic research, the special needs for managing a basic research program were not tended to either. A system called "Research Planning Objectives" (RPO's ) was instituted to require that research meet predetermined objectives. (35) Also an important aspect of the ability to function effectively was determined by the status of the institution within the Air Force. While the Air Force research organization was seven layers below the Secretary of the Air Force, a comparable office in the Navy, the Office of Naval Research was highly placed in the Navy reporting just below the Under Secretary of the Navy. (36) The Navy office could function more effectively than the comparable one in the Air Force because of its more favorable position in the administrative hierarchy. By the late 1950's, basic research in the Air Force was thoroughly discouraged. A number of factors had taken their toll, including the need to provide for relevance to military objectives, which Kumons calls "camouflage." Contracts with researchers and their universities were short term and resulted in severe cutbacks and even in the possibility of the Air Force having to default on its contracts in the budget crunch which occurred in Eisenhower's administration just before the Soviet Union's launch of Sputnik. The categories of research allowed within the RPO's were also stifling to the researchers. The staff of the Air Force Research Agency by the late 1950s was thoroughly demoralized. The events that followed the launch of Sputnik, however, renewed a recognition that there was the need for basic research. Through a series of administrative changes, the Air Force research organization was moved up in the administrative structure. It was then only one step below the head office. The need to head the agency with a civilian scientist was recognized and Dr. Knox Millsaps, an applied mathematician, was appointed. The policies that Millsaps instituted in the agency, which by this time had its name changed to the Air Force Office of Scientific Research (AFOSR), led to significant changes in the ability of the office to support basic research and researchers. Millsaps decided that pure research had to be differentiated from applications. He understood that the business of the Air Force was to create new weapons, but that this end product was different from the research that was necessary to make this possible. Pure research had to be recognized as necessary in its own right. In place of required progress reports, researchers supported by AFOSR were encouraged to publish in science and engineering journals. Millsaps reasoned that what good was a report from a contractor "if the report's findings weren't acceptable for publication in a scientific journal?" (37) Researchers were supported based on the quality of their work, not on the nature of the project that was proposed. Even if two researchers were studying the same topic, there was no reason not to support them as they would both contribute to the field. The only reason not to support a researcher was if the person wasn't top rate, and that was in general not hard to detect before giving support, considering the open communication that was common in the sciences. Another major problem that Millsaps recognized was the length of research contracts. In other government matters, contracting on a short term basis was a way to protect the Air Force so that the product was delivered before it was obsolete, etc. In research, however, this wasn't true. A short term contract there prevented the director of a research program from properly planning a program. Instead short term objectives were substituted. Millsaps committed himself to lengthening the period of time for contracts. Also he realized that it was necessary to do away with the RPO criteria as the applied research categories did not accurately portray the research that was being done. Even worse these categories excluded certain areas of research. Instead scientific fields were chosen to categorize the research areas that AFOSR would support. Millsaps decided he would present the research as it was being done to Congress, and ask for support on its merit. He recognized that camouflaging the nature of basic research for reporting purposes to Congress carried with it the danger of ending the basic research. That substituting semantic changes in the nature of the kind of research to get it funded was, at best, an effort to create protective camouflage, "and at worst they symbolize a petard of intellectual dishonesty on which sooner or later some ill-disposed official at higher levels will hoist the Air Force basic research program."(38) Also Millsaps believed that those doing basic research should not be connected to or required to transition their research to the applications people in the Air Force Laboratory. He reasoned that the research supported by the Air Force was only a small portion of the pure research in areas of interest to the Air Force that was being carried out. Far better, he believed, to hold those working in the Laboratory responsible for creating applications that are not limited by the work done by the researchers supported by AFOSR. That the laboratory scientists should on their own examine all the scientific literature and determine how it would meet their needs for applications. Millsaps also raised the government employee ranking of the program managers. And he spent $100,000 to fix up the headquarters to improve the work environment. 3. The Office of Naval Research A similar set of events marked the experience of the Office of Naval Research (ONR). However, this agency had been set up through a congressional mandate. Formed by a Congressional order, it was in a stronger position than the Air Force research agency. Harvey Sapolsky, in his book Science and the Navy (New Jersey, 1990) notes how legislation setting up the Office of Naval Research made its way through Congress in 1946. Public Law 588, creating the ONR, "not only gave it equal status with the Navy's materiel bureaus," Sapolsky notes, "but also gave it independence from the fleet and the growing authority of the Chief of Naval Operations." (39) This placement is important, and Sapolsky describes how in 1946, the Navy was a "bilinear organization", the two parts of which were the "fleet itself" and the "shore-based materiel support establishment", both of which reported separately and on an equal status to the Secretary of Defense. (40) The ONR was headed by a naval officer, as required by the law establishing it, who was the Chief of Naval Research. But it was established to report to the Secretary of the Navy in an equal status to the fleet and the materiel support establishment. "This formal independence from the uniformed Navy," Sapolsky writes, "was an advantage in that ONR was neither absorbed by the routine demands of the fleet nor subject to the reorganization whims of senior naval officers. But ONR was also somewhat isolated from the main interests of the Navy - the current and future health of naval forces." (41) And this latter situation could lead to the Navy turning "hostile" to the Agency, Sapolsky recognizes. The ONR hired civilian staff members, and early in its development, the control of the office shifted from the military officers to the civilian staff. (42) Sapolsky describes some of the politics of the delay in the legislation supporting the creation of the NSF until 1950, and then the efforts of the ONR to manuever so that they wouldn't be disbanded when the NSF was established. (43) The initial funding for the NSF was much less than expected and the ONR maintained its right to continue in existence. With the US entry into the Korean war, however, pressure was brought to bare on the ONR to redirect its activities toward more applied research that would be in the words of its critics, "relevant" to the Navy. "Despite its sophistication, ONR could not totally resist in the pressure", writes Sapolsky, pointing out that money "became available for weapon-related projects, but not for the basic research that the ONR had been funding." (44) Congress had initially funded research and development as part of the procurement appropriation to the Navy. In FY 1954 budget, Congress established a category "Research and Development Navy" which was to contain all the R & D appropriations for the Navy, under the context that they didn't want the Navy to shift monies from R&D to weapon procurement, and vice versa. (45) The Chief of Naval Research was subsequently given responsibility for coordinating all of the Navy R&D programs. Sapolsky notes how this new responsibility brought the ONR 'directly into the center stage of naval politics'. (46) He explains that the Navy was organized around "competing technologies" like "aircraft," "surface ships" and "submarines" and that the selection of weapon systems for procurement directly affects how power is distributed within the Navy. The new mission, therefore, put the Chief of Naval Research into the position of deciding the future of power within the Navy, which was too political a position for someone with the limited authority of a directive from the Secretary of the Navy that the Chief of Naval Research had been given as its authority. (47) The new mandate was soon removed. But the incident demonstrates the internal Service politics that the civilian scientists within the Department of Defense were facing with their mandate to provide for future scientific and technological directions for the Services. The experience of the ONR during the 1950's was one of continuing pressure to support applied research rather than basic research. Internally there was an effort to resist the pressure. Sputnik brought new challenges in that the Advanced Research Projects Agency (ARPA) was created at the level of the Office of the Secretary of Defense, along with a trend to consolidate research. (48) In the 1960s, there was continual pressure against the basic research that the ONR wanted to fund, and with more centralized decision making being carried out within the Department of Defense, the ONR was continually required to apply cost effective analysis to its research grants, to demonstrate the future expected benefits that would come to the Navy. (49) By the end of the 1960's the Vietnam War meant that military budgets were increasing. Along with these increased expenditures came increased pressure for relevance related research. In 1969 Senator William Fulbright, Chairman of the Foreign Relations Committee (50): "proposed an amendment to the FY 1970 Military Procurement Authorization Act, Section 203, which stated: `None of the funds authorized to be appropriated by this Act may be used to carry out any research project or study unless such project or study has a direct or apparent relationship to a specific military function or operation'." Sapolsky describes the demoralizing effect this had within the ONR and the wedge it created between the military and civilians within the Navy. Describing the differences between the military and civilian scientific staff within the Navy, he writes(51): Naval officers tend to be advocates for the Navy's short- term needs even if they hold positions where the opposite perspective is more appropriate. They are, after all, the organizational members who will do the fighting if combat does occur. Moreover, they are subject to a personnel system that encourages them to press for immediate results. They are given assignments of short duration--two, three, or four years--and their promotional opportunities depend in large part on fitness reports that describe their accomplishments during these brief tours of duty. Because of their constantly changing assignments, they only rarely develop the deep expertise required for scientific standing. The Navy's civilian scientists tend to have the opposite bias. They are taught and most ardently believe, that all scientific knowledge is useful even though the most important applications of this knoweldge may not be revealed for years. They know that significant scientific results are often painstakingly acquired. And they all arrive with, and are permitted to develop further expert knowledge in subdisciplines whose utility to the Navy they learn becomes vital and is appreciated at only unpredictable intervals. It is not surprising that these perspectives clash. The continuing pressure on the ONR staff, Sapolsky demonstrates, resulted in the fact that though "bureaucratic subterfuge saved the basic research program for a time," more permanent redirection into an applied research program was the inevitable result. (52) "Good science still gets done, but often in disguise," he observed. (53) The experiences of these two agencies within the Services help to set a basis for examining what happened with the Information Processing Techniques Office (IPTO) set up inside of ARPA in 1962. (to be continued) Footnotes (Footnotes for 1-3 will be for an introduction that will be part of the final paper. Hence the footnotes here start at 4) (4) Daniel J. Kevles, "Scientists, the Military, and the Control of Postwar Defense Research: The Case of the Research Board for National Security: 1944-46" in Technology and Culture, vol 16 no. 1, Jnauary 1975, pg. 24. Kevles refers to reference to this conference in "Navy Department Conference to Consider Needs for Postwar Research and Development for the Army and Navy," April 26, 1944, OSR, Director's Files, Postwar Planning Reports. (5) Ibid. p. 24 (6) The National Academy of Science (NAS) was created as a private organization by an Act of Congress on March 3, 1963 "in order better to focus scientific talent in the Civil War." The act described the function of the Academy as "whenever called upon by any department of the Government investigate, examine, experiment, and report upon any subject of scinece or art." (See Stewart, pg. 4) During WWI, President Wilson requested that the NAS expand and create a National Research Council (NRC). The Executive Order No 2859 of May 11, 1918 authorized the NAS to create the NRC. (7) Ibid., pg. 27. (8) Ibid. pg. 29. (9) Ibid., pg. 33. (10) Ibid., pg. 36. (11) Ibid., One of the reasons given in a memorandum on this issue cited by Kveles is interesting. He quotes a memorandum form Arnold Miles to Donald C. Stone, February 13, 1945, saying "The more broadly the armed services conceive of warfare, the more intent they will be to keep scientific discoveries of all kinds under cover. Is the Board -- composed of representatives ofthe armed services and of scientists who wuld undoubtedly represent scientific interest (under no conditions to be confused with the public interest)--capable of representing the public interest in this very difficult area? I think the answer is obvious -- emphatically no." (12) Ibid., pg.39. (13) Ibid., pg 43. (14) Ibid., pg. 43-44. (15) Ibid., pg 44. (16) Nick A. Kumons, Science and the Air Force, Arlington, VA, 1966, pg. 1. (17) Ibid., pg. 1. (18) Ibid., pg. 2. (19) Ibid., pg 5. (20) Ibid. (21) Ibid., pg 10. (22) Ibid., pg 12. (23) Ibid., pg 15. (24) Ibid., pg. 18. (25) Col Oliver G. Haywood, Jr. "The Air Research and Development Program," Journal of Engineering Education, XLIII (March 1953), in Kumons, pg 26. (26) Ibid. Col. William O. Davis, transcript of personal interview with Dr. Ernest Schwieber, 8 November 1956, pg. 27. (27) Ibid., pg 32. (28) Ibid., pg 57. (29) Ibid., pg. 59. (30) Ibid., pg. 61. Also see "Haywood's belief, that AFOSR had a unique mission which could be properly performed only if the organization was uniquely placed in the Air Force's R&D hierarchy, was still the common belief among AFOSR's staff." (pg 67) (31) Ibid., History of the Air Research and Development Command, 1 Janaury - 30 June 1954 (ARDC Historical Deivsion) I, 155 (herein cited as History of ARDC) in Kumons, pg. 68. (32) Ibid., pg. 69-70. (33) Ibid., pg. 70 (34) Ibid. "What was once cant became gospel."(pg 70) (35) Ibid., pg 82. (36) Ibid., pg. 85. (37) Ibid., pg. 147-148. (38) Ibid., pg 148. (39) Sapolsky, pg. 47. (40) Ibid. (41) Ibid. (42) Ibid. pg. 49 (43) Ibid. pg. 55. (44) Ibid., pg 64. (45) Ibid., pg 65. (46) Ibid., pg. 66. (47) Ibid., pg. 67-68. (48) Ibid., pg. 72. (49) Ibid., pg. 72-73. (50) Ibid., pg. 43-74. (51) Ibid., pg. 77-78. (52) Ibid., pg. 81. (53) Ibid., pg. 80. The following notes will be part of an introductory part of the paper about methodology: (1) Among those recognizing the essential contribution to the postwar development of computer science in the US and in the world by the ARPA/IPTO community and leadership include NSF personnel from this period, European computer scientists like Barber and Pouzin, and ARPA documents after the office was ended. For example. Fred Weingarten of the NSF in an interview done by 26 Sept 1990 by the Babbage Institute says: "DARPA really was the formative agency for computer science. I mean, they weren't the first. ONR probably was the first, but they really swamped with their money and the centers of excellence." (2) See for example "Advisory Committee on Industrial Innovation Final Report" United States Dept of Commerce, Washington DC, September 1979. It describes the need to study previous examples of successful government activity to determine how to learn from those toward creating needed government institutional processes or entities. Discussing the need to learn from previous experience in supporting innovative development, they proposed that "The National Bureau of Standards should prepare a paper on the experience of the Energy-Related Invention Evaluation Program to date, including followup on successful inventions, recommendations for program improvement, and opinions on the expandability of the program to areas other than energy. Other government experience in this vein, such as that of the National Transportation Safety Board, should be surveyed and evaluated." They also proposed studying the experience of the Office of Naval Research to learn from its experience. Another example of recognizing the need to study previous experience in government toward determining the lessons toward future activity is included in the National Research Council report "Funding A Revolution: Government Support for Continuing Research" National Academy Press, Washington, 1999. The report says: "As the computer revolution continues...the federal government continues to play a major role, especially by funding research. Given the successful history of federal involvement, several questions arise: Are there lessons to be drawn from past successes that can inform future policy making in the area? What future roles might the government play in sustaining the information revolution and helping to initiate other technological developments. This report reviews the history of innovation in computing (and related communications technologies) to elucidate the role the federal government has played by funding computing research and to identify factors that have contributed to the nation's success in this field." The report goes on to describe how "Historical analysis is one means of informing debates over the role of the federal government in computing research. History provides empirical evidence of the success and failure of different policies over time and it offers evidence from which patterns can be seen and conclusions drawn about the funding process in particular and innovation in general....Case studies are a standard tool of historical analysis, allowing one to move more deeply into the mix of events, people, and organizations associated with the funding of computer research. Case studies provide an intimacy with history akin to that experienced by persons who lived it. They present the messy details of real-life experiences not available in abstract, quantitative analysis." (pg 18) Though the study notes the limitations of such studies, it refers to Richard Nelson's research. "Nevertheless, he noted, it is possible to make judgments about the kinds of policies that are feasible and effective in different contexts." (See Richard Nelson, ed. "Government and Technical Progress: A Cross Industry Analysis", Pergamon Press, 1982, p. 454) (3) The OSRD was created as a result of the sense among scientists that the world and science had changed and that therefore a new form of scientific research and development organization would be needed by the U.S. government to effectively utilize science and technology in the war effort. While previously, military officers could be depended on to know the scientific and technical needs of the Services, Irvin Stewart explains in Organizing Scientific Research for War, modern science meant that that was no longer true. He writes: Modern science has progressed to the point where the military chieftains were not sufficiently acquainted with its possibilities to know for what they might ask with a reasonable expectation that it could be developed. The times called for a reversal of the situation, namely letting men who knew the latest advances in science become more familiar with the needs of the military in order that they might tell the military what was possible in science so that together they might access what should be done." Irvin Stewart,Organizing Scientific Research for War NY, 1980, pg. 6. Stewart describes how the National Defense Research Committee (NDRC) was initially created as part of the Council of National Defense by an order of the Council with President Roosevelt's approval on June 6, 1940. But then to meet additional needs that the NDRC couldn't meet, an Executive Order of the President was issued on June 28, 1941 creating the Office of Scientific Research and Development. (See page 35-36) The OSRD was placed in the Office for Emergency Management of the Office of the President. * for part I see "Computer Science and Government: ARPA/IPTO (1962-1986) Creating the Needed Interface" http://www.columbia.edu/~rh120/other/arpa_ipto.txt Last updated January 27, 2000 part I http://www.columbia.edu/~rh120/other/arpa_ipto.txt part II http://www.columbia.edu/~rh120/other/basicresearch.txt part III http://www.columbia.edu/~rh120/other/centers-excellence.txt part IV http://www.columbia.edu/~rh120/other/computer-communications.txt part V http://www.columbia.edu/~rh120/other/birth_internet.txt part VI http://www.columbia.edu/~rh120/other/birth_tcp.txt