their eventual impact on society will be an important step in the evolution of the science organization at the presidential level. Such analysis must be carried out without the initial constraints of agency jurisdiction, and in intimate relationship with the decision and policy-making processes in the Executive Office of the President.

It is clear that we need more systems analysis on a government-wide scale. I do not mean the formal and sometimes sterile approach of professional systems analysts. Rather, I refer to analysis that is both tied to the decision-making function and involves the creative thinking of a large number of people looking at the inventive process, without undue concentration on the techniques and methodology of systems analysis.

There are many basic questions facing the government that we have been unable to analyze in a systematic way-questions like these: How many graduate schools, of what kind, does the country need to meet its present and future needs? What is the effect of the various development programs on the future requirements for research support? What trade-offs were we really making when we initiated a space program, and what trade-offs will we be making if we cut it back now? We need studies like these as part of a continuing assessment program. The Office of Science and Technology could eventually evolve into an office of planning, evaluation, and analysis, looking broadly at national problems with some scientific or technological component but extending well beyond the purely technical areas. The OST has been moving in this direction in its work in environmental pollution, urban needs, and the world food problem. In fact, OST has a newly formed Office of Energy Policy Coordination which is undertaking a broad study of the many important energy policy questions that affect more than one agency of government. The new policy questions are occasioned by the rapid advance of nuclear power on the energy scene; the need to reconcile air-pollution and water-pollution programs with our demand for low-cost energy; the growing demand for energy in relation to available supplies at economic prices, especially supplies of natural gas; the basic question of future import policy concerning oil, gas, and uranium; government policy toward developing sources of oil and gas from shale and coal to supplement natural supplies; and many others.

As I have indicated, an essential feature of such studies is that they be carried out in such a way that they are an integral part of the policy-making process; that they deal with the real world economic and political constraints, without accepting them as immutable.

Such an evaluation capability should be part of the forward planning function that needs more explicit recognition in the Executive Office of the President. By "planning" I do not mean a rigid blueprinting of the future. Rather, I mean a best current projection of the future, and of alternative futures, based on present activities and planned new ones. We simply are not smart enough to put together large-scale plans for many things at the present time. However, by developing a capability for analysis, it should be possible more and more to chart the future analytically rather than through mere intuition and debate. There are some things that cannot be done without large-scale planning. A national telephone system required an overall plan, and systems analysis and engineering were needed to put it together. We have undoubtedly foregone some competition in the process, and perhaps some of the components are more expensive than need be, but the need to eliminate internal incompatibilities was overriding. Similarly, despite the political fragmentation of many communities, water and sewer systems must be put together according to a plan. For largeweapon or space-systems development, the complexity of the many efforts which must jell, with a lead time of 5 years or more, requires a working plan. Many more big national problems are forcing us in this direction. The structure of university science may well be approaching that divide where the need for overall systems planning will take precedence over the goal of obtaining maximum health of each of the parts taken one at a time.

Undoubtedly, we will have to face up to the need for more comprehensive planning. We can begin-in fact we already have begun to isolate those manageable pieces of the larger problems that lend themselves to analysis, and, as further areas yield to analysis and we better understand the boundary conditions, it should be increasingly possible to predict likely outcomes from given actions. Of course, as scientists we recognize that the best of analysis cannot predict the outcome if we do not know the relevant inputs, or, as is so often the case in complex problems, when we are not even sure that we know what all the relevant

variables are. In such situations we rely on the carefully controlled and evaluated experiment. The experiment is the lifeblood of science, and we must learn to use it effectively in other areas. For example, in dealing with urban problems we must learn to employ experiments to help answer the larger questions that do not yield to analysis.

We shall have to foster many experiments involving large systems, but naturally we need to know how we will evaluate them when they are finished. Rational analysis coupled with experimentation should make it clearer what we need to do by experiment and what choices are available through analysis. Unfortunately, we have too often substituted bureaucratic and political processes for either rational planning or experimentation. In a democracy this may always be the case, but the analysis will at least provide a better basis for political discussion.

A second principle of government policy ought to be to maintain competition. Insofar as government actions and organization are concerned, many people now suggest a highly planned economy for science, with a rigid separation of functions and a careful elimination of duplication. Our successful experience suggests a contrary course. Most government agencies that have remained virile and avoided deterioration have done so, in part, by stepping on each other's feet. As a general rule, if there is a large opportunity or need at stake, it is profitable and appropriate to employ both competition and careful planning.

More importantly, basic science is both a cooperative and a highly competitive activity. Its progress depends on the stimulation provided by competition. For a vivid illustration I refer you to James Watson's fascinating book The Double Helix (3). In science, as in economic processes, competition stimulates the quality of performance and must be fostered, together with the cooperation which comes through an open, widespread, and effective communication system among scientists.

PROPOSALS

Finally, I want to make some specific proposals.

First, I believe the Office of the Science Advisor needs strengthening, not only through more staff capability for analysis and planning but through the addition of full-time top-level people. In short, I propose that the Science Advisor be made the head of a three- to five-man Council of Scientific and Technical Advisors. My reason is simple-the range of matters he must at present consider is so broad and his responsibilities are so extensive that he needs help. Alternatively, one might add three assistant directors to the present director and deputy director.

I also believe that, provided the staff resources were available, it would be wise to ask such a council to submit to the President and the Congress an annual report on the state of U.S. science and technology, roughly analogous to the annual Economic Report.

Second, we should reexamine a possibility we put aside some years ago— namely, that those scientific activities not tied to the central purposes of an agency be considered for inclusion in a department of science, with the National Science Foundation as a core. Science has now assumed such importance to the nation that its position would be stronger if it had a voice at the Cabinet table. However, in making that proposal I want to make it clear that I would not consider concentrating all of our science activities into a central agency. A strength of the American establishment is the realization that science is part of everything. Those research activities which are integral to a department's mission or which form the basis for its future should be left where they are. More than that, agencies should be encouraged to strengthen their research and development base. But there are other scientific activities of agencies which may be somewhat peripheral to the main job of an agency but are nonetheless important, and these would flourish if transferred to a department of science.

In determining the organizational elements of a department of science, thought will have to be given to the department's relationship to advanced education on the one hand and technological advance on the other. The more the department is oriented toward new technology, the less it is equipped to deal with academic science and advanced education, including the humanities. The more it is oriented toward basic research and academic science, the more it is fitted for a broader role in higher education. On this score, one could invent several cuts that would represent an improvement over the present situation, but I am far from sure what the best cut would be. My present feeling, though, is that the critical questions concern basic research and higher education, and that technological development is more appropriately conducted by agencies with specific tasks and missions.

In the power equation of Washington, such a department of science, if it is to be influential, should have a budget of $2 billion or more. Its principal officer would have line responsibility and public accountability and, most importantly, the interest and confidence of the President, the attention of the Bureau of the Budget, and the ear of the Congress.

With a strong cabinet officer for science in the Executive Branch, there would automatically be a strong congressional counterpart committee having a broad interest in the problems of science and technology, not a minor or incidental interest. We already have committees like the Joint Committee on Atomic Energy, and the House Science and Astronautics Committee that are broadly educated in particular spheres of scientific and technological activities, and I am confident we could have committees of this caliber to supervise this department too.

CONCLUSION

Both the problems and opportunities facing government science policies loom larger than ever before us. I have been privileged to have had a part in setting U.S. science policy and am proud of what has been accomplished so far.

Despite the last 25 years' evolution of the U.S. science structure in the U.S. government, we are still in the early stages of learning how to realize the potential of science and technology for the national good. But we have built a strong foundation, on which further additions and structural changes can be made with confidence.

REFERENCES

1. V. Bush, Science: The Endless Frontier (Government Printing Office, Washington, D.C. 1945).

2. G. Soule, Economic Forces in American History (Sloane, New York, 1952). 3. J. D. Watson, The Double Helix (Atheneum, New York, 1968).

[Reprinted from Science, July 4, 1969]

DO WE NEED A DEPARTMENT OF SCIENCE AND TECHNOLOGY?

(By Herbert Roback)

A RECURRING ISSUE GETS A FRESH APPRAISAL AND A POSITIVE ANSWER

1

Departmental status for science in government is not a novel idea. It was broached 85 years ago by a committee of the National Academy of Sciences reporting on the organization of government science bureaus. The committee was appointed in 1884 at the behest of a member of the National Academy, Theodore Lyman, who was also by unique coincidence a member of Congress. The scientist-congressman, whose National Academy standing was gained by his researches on the Ophiurida, was instrumental in placing a rider on a sundry civil appropriation bill which set up a joint congressional committee (called a commission) to study the organization of the government science agencies (1). This was a compromise measure, Lyman told his House colleagues in urging its acceptance, his main concern being to save the Coast and Geodetic Survey from takeover by the Navy (2). Apparent duplication between the Coast Survey and the Navy's Hydrographic Office in charting coastal waters had led the Navy to espouse merger, a proposal which had considerable appeal in the 48th Congress. The legislative rider directed an investigation of the activities and interrelationships of the Coast and Geodetic Survey, the Hydrographic Office of the Navy, the Geological Survey, and the Signal (Weather) Service. The National Academy committee, enlisted by Lyman, gave technical support to the congressional commission.

Men of science were leery then, as they are now, of military dominance in scientific enterprises. Many of them argued that science agencies should be taken from, rather than placed in, the military departments. For example, they wanted the Naval Observatory to be a national observatory, and the weather

1 The author is staff administrator of the Military Operations Subcommittee, Committee on Government Operations, U.S. House of Representatives. This article is adapted from notes prepared for the Symposium on Science and Engineering Policies in Transition, Carnegie Institution of Washington, 18 and 19 December 1968.

service to be removed from the Army Signal Corps. At the same time, they recognized that better coordination of the government's scientific work was needed, and various proposals were made toward that end. The aforementioned report of the National Academy committee crystallized the issues. This group was convinced that the science agencies should be pulled together "under one central authority," but the particular form of organization they left to the future and to Congress. Then the committee ventured this cautious but significant observation (3).

. . The best form would be, perhaps, the establishment of a Department of Science, the head of which should be an administrator familiar with scientific affairs, but not necessarily an investigator in any specific branch. Your committee states only the general sentiment and wish of men of science, when it says that its members believe the time is near when the country will demand the institution of a branch of the executive Government devoted especially to the direction and control of all the purely scientific work of the Government."

The NAS committee went on to say that, if public opinion was not yet ready to accept a Department of Science, the next best step would be to move the several scientific bureaus into one of the existing departments. Even then coordination would not be automatically insured, in the committee's view, and so they recommended the "organization of a permanent commission to prescribe the general policy for each of these bureaus." The commission would "examine, improve, and approve" plans of work and expenditures and recommend efficiency measures but abstain from administrative involvement. This would be a ninemember commission composed of scientists drawn from government and private life (4).

END OF AN IDEA

The congressional commission, reporting in 1886, gave short shrift to the suggestions both for a Department of Science and a supervisory commission. A new department was held not justified by the degree of duplication in existing scientific agencies; a coordinating policy group was deemed impracticable because department heads could not very well relinquish to subordinates and outsiders their responsibilities for general direction and control (5). With this dismissal by an agency of the Congress, the Department of Science idea died aborning, though it was actively debated at the time in scientific circles (6). In the ensuing decades not much was heard about it. Proposals for government departments were made in the fields of health, education, labor, industry, commerce, and agriculture, separately or in various combinations, and three cabinet departments (Agriculture, Commerce, Labor) were established between 1885 and 1945. Not until 1946 was the Department of Science idea revived, at least in the legislative halls. Clare Booth Luce, then a Representative from Connecticut, introduced a bill (H.R. 5332, 79th Congress) to create a Department of Science and Research, stressing the need for national self-preservation in the atomic age and the importance of attracting young people to science careers. Mrs. Luce said: "Only the prestige which attaches to a regular member of the cabinet will render the findings of any scientific body of sufficient weight to command the constant attention of the highest officials of the Government in the consideration and formulation of policy" (7). The bill was pigeonholed by a House committee.

Vannevar Bush was working for the establishment of an independent agency, which he called the National Research Foundation, to sponsor research of military as well as civilian interest (8). He proposed that it be governed by a director and part-time board of nongovernment scientists. A separate group of nongovernment scientists, which he called a Science Advisory Board, would coordinate the work of government science agencies. These proposals, outlined in Bush's 1945 report to President Truman, "Science-the Endless Frontier," were modified in legislative measures to become a National Science Foundation. The bill which the Congress passed was vetoed in July 1947 by President Truman, who objected to control of government science policy by an outside board (9). The criticisms were reminiscent, in some respects, of those heard in 1884-5 to the National Academy committee's proposal for a science policy commission which would include outside as well as government scientists.

The Bush report was followed in 1947 by the Steelman report, "Science and Public Policy," which went over much of the same ground but with closer orientation to the routines of governmental administration. The Steelman report called for a National Science Foundation to be organized "on sound lines" and suggested that the agency be located in the Executive Office of the President until other federal programs in support of higher education were established,

after which time consideration could be given to grouping all such activities, including the National Science Foundation, in a single agency. The Steelman report also favored a part-time governing board for the NSF, but government as well as outside scientists were to be included. It also recommended the creation of an interdepartmental committee on scientific research and development, a special unit in the Bureau of the Budget to review government science programs, and a member of the White House staff to be designated by the President for purposes of scientific liaison (10). The Steelman report eschewed any radical department from the existing framework, presumably meaning that a Department of Science was not in the cards. Three years elapsed, however, before the differences in the several approaches to a National Science Foundation were compromised and a bill finally enacted into law (11).

LEGISLATIVE RENEWAL

Sputnik generated a new debate on departmental status for science in the Congress led by Senator Hubert H. Humphrey. On 27 January 1958 a broadbased bill, S. 3126, was jointly introduced by Senators Humphrey, McClellan, and Yarborough to create a Department of Science and Technology which would coordinate and improve federal functions relating to the gathering, retrieval, and dissemination of scientific information; provide educational loans to students in certain science fields; establish national institutes of scientific research; and establish cooperative programs abroad for collecting, translating, and distributing scientific and technological information. A day later Senator Kefauver introduced S. 3180 to create a Department of Science. Both bills were referred to the Committee on Government Operations.

Jurisdictional questions were raised, presumably because the bills went beyond organizational matters into policy, and at the request of Senator Lyndon B. Johnson, they were referred anew to his Senate Special Committee on Space and Astronautics, which had been created to consider the government's response to the Russian triumph in space. Without these bills, the Committee on Government Operations was unable to hold hearings in the 85th Congress on the proposal to establish a Department of Science and Technology, which was incorporated in Title I of the bill, but it directed its committee staff to maintain a continuing study of that area. The Humphrey subcommittee did manage, after an agreement reached with Senator Johnson, to hold some hearings in May-June 1958 on a limited aspect of Title I, the proposal for a scientific information center (12). To narrow the jurisdictional issue and regain control of the organizational aspect, the sponsors of the Humphrey bill, now reinforced by Senators Ervin, Gruening, and Muskie, split off Title I and introduced it, with certain revisions, as S. 676 in the 86th Congress. It proposed a transfer to the new department of the National Science Foundation, Atomic Energy Commission, National Aeronautics and Space Administration, National Bureau of Standards, and certain activities of the Smithsonian Institution. By then the impetus for a new department was considerably diminished by NASA's presence. The thrust of science organization was less to coordinate and align than to reach out and do, for Sputnik had caused hurt pride and fear in the nation. It was difficult to make a case for legislating a new department to absorb NASA when the ink was hardly dry on the President's signature to the National Aeronautics and Space Act (13). Indeed, the rush of legislative events and the flurry of organizational activity in the executive branch during 1958 outpaced the committee's deliberations on the suitable form of a bill. The Congress created along with NASA an Aeronautics and Space Council and a standing committee in each house to monitor space and related activities. The Defense Education Act gave support to science education and facilities. A reorganization act for the Department of Defense established a Directorate for Defense Research and Engineering. The Advanced Research Projects Agency, previously established as the military's own response to Sputnik, was made an adjunct of the new directorate. The President acquired a Special Assistant for Science and Technology and gave White House status to the Science Advisory Committee. The Federal Council for Science and Technology replaced a looser interdepartmental committee of similar function. Science advisers were assigned to both the Secretary of State and the Secretary General of NATO. A NATO science committee signified the outward reach of science for defense, while "Atoms for Peace" and the International Geophysical Year represented a peaceful gesture to a world community of science. "Altogether," as James R. Killian, Jr., said before the AAAS in summing up government sci