Page images

The Committee report uses the word "science" to cover a multitude of activities including basic research, routine engineering, data gathering, and engineering for production—all under one title. But the different activities included in "science” require different approaches in their management. Therefore, one point I wish to make is that the question as posed by the Committee is too broad. If we wish to respond to the problems which prompted the question in the first place, we ought to distinguish among basic research, developmental studies, developmental engineering, demonstration projects, and engineering for production. Each activity requires its own managerial style.

Science and technology are not ends in themselves. Managerial styles should not be chosen just to make science "healthy." They should be chosen to make science and technology responsive to human needs.

As an idea moves from research to development and then to production it needs to be managed differently.

The research stages are the least expensive and the most resistant to close management. Bob Charpie, former president of Bell and Howell, made the point very well when he said, "the kinds of revolution, the kinds of upsets, the kinds of dramatic change that result in consequential action on the national economy often come out of the blue from unexpected quarters. The photographic industry did not invent instant photography; the textile industry did not innovate synthetic textiles; and tungsten carbide did not come out of the machine tool industry. In fact, a tabulation of 180 major technical innovations made in the first six decades of this century, shows only 20 percent of them came from within the industries, which they finally became part of. It's not only humbling, it is positively shocking to see how many of these came from outside the establishment."

He also said, "There are 200 enterprises in the U.S. which spend 94 percent of all the R&D dollars, yet 70 percent of the important innovations came from outside these 200 companies in the first sixty years of the century."

Professor Joseph Ermenc of the Thayer School of Engineering at Dartmouth College has made tape recorded interviews with important inventors, whose contributions have had a strong impact on our economy. Listening to Carlson describe the development of Xerox, or to Sikorsky talking about the early days of the helicopter, or Pilkington explaining his difficulties in getting anyone to accept his float method of making plate glass, I cannot help but wonder how it is that any invention survives and sees its way to the market place. As far as I can tell, just about every method we devise to attempt to control and promote invention, in fact serves to impede the inventor. Sometimes when we see a blade of grass growing up through a crack in the city sidewalk, we marvel at the tenacity of nature, which some how or another overcomes all of man's obstacles to growth. I feel the same way about inventors and inventions.

At the present time I happen to be deeply involved in planning for the development of weather modification and I, therefore, would like to cite it as another example of the relation between research and management. I was intimately associated with the earliest activities in this field. I was assigned to the Army Air Corps at Wright Field, during World War II and put to work on problems of ice prevention in aircraft. In this way, I met Irving Langmuir, who was then working on the problem of radio interference from static electricity on aircraft. His unorthodox approach to this problem got him into a quarrel with the authorities then in the field. These authorities controlled his budget. I was present in 1943 when they decided that Langmuir's project would be denied further financial support. I was so impressed with Langmuir that my office gave him a new contract to look into the trajectories of cloud droplets streaming past airfoils, thereby helping us to understand how ice forms on an airplane wing and to design equipment to protect against the ice. But Langmuir didn't confine himself to the question posed in the contract. He wanted also to know why the drops of water turned out to have the sizes they did. This led him to investigate the conditions of cloud growth on the summit of Mt. Washington.

When Langmuir began to understand the details of cloud formation, he turned his attention to the possibilities of cloud dissipation. Since this work was moving outside the field of the original contract, I decided to take his ideas to the Weather Service, the place in which the support of weather research would logically have been centralized. The reaction of those then in charge was so negative that I was simply laughed out of the place. My belief in Langmuir

1 "Science and Technology," November 1968, p. 64.

was strong, however, and I therefore, arranged to support the work. Within a year, the first successful cloud seeding was accomplished. Today the Federal Government spends approximately $11 million per annum trying to develop a workable system, based on the research which began with Irving Langmuir. The point I make is that if science had been strongly centralized in those days, we might never have reached the point of demonstration. Langmuir summed it up in the statement, "The hardest thing to sell is a new good idea. If it is good and it is new many people will have to change their minds. They simply won't be willing to do so." I hasten to add that today the Weather Bureau, and all of ESSA, are staunch supporters of experiments in weather modification and wholeheartedly favor the work.

Almost without exception, every significant new idea in science has had to fight its way against the entrenched establishment. Scientists are no different than other people. In spite of the public image which portrays scientists as cold, logical, reasoning machines, I can assure you that the scientific establishment often resists change for much the same reasons as other sectors of society. The introduction of new ideas usually means the introduction of new rules of play and someone who is on top will certainly have to step aside in the readjustment.

My conclusion from a study of history, and from my own involvement in scientific research, is that those who think they can centralize the control of research ought to forget it and for two good reasons :

(1) They cannot succeed.

(2) If they even meet with partial success the major result will be the choking off of creativity. The things which can be done to enhance the probability of useful research are mostly in the categories of avoiding those things which hinder the process. Good scientific research requires a permissive atmosphere, reasonable financial support, and above all freedom. Freedom to try, freedom to publish, freedom to persuade, freedom to travel, freedom to make mistakes, and freedom to contribute.

To put it more bluntly, one should not try to centralize the control of research because it cannot be done without either serious detriment to the results or loss of necessary opportunity for the man with the new idea.

I have dwelt on the topic of research because I think it is important enough to justify a separate discussion of why we ought not to centralize it. On the other hand, I believe that engineering development can be managed and can be controlled much better than it is. If we consider the relative amounts of money spent on research and on development, it becomes obvious that the biggest gains to the taxpayers are most likely to come through attention to the processes of development and preparation for production. Dr. Hillier of RCA has suggested that as a crude measure of the funding required, we can say that for every dollar spent in research on a project, $10 must be spent in development and $100 in production. The actual outlay of funding in research, development and production does not correspond to the ratios 1-10–100, because many projects in research do not lead to velopment. Some of the developmental projects do not lead to production. But for any one invention the ratio 1-10–100 does seem reasonable. At the research end of the scale, we seldom know which idea will turn out to be the one we eventually wish to push through to development. That's why we start so many projects in research. Attempts to impose strong controls on research are apt to be counter productive. But when the decision has been made to move a research idea forward, and into production, the development ought to be managed very well. Poor management not only introduces costly mistakes on a large scale, but it can also delay the day when an idea reaches production. Revenue lost through tardiness never shows up on the balance sheet, but it is a real cost.

Earlier in this testimony, I suggested that we ought to consider the problems which generated the original question, rather than the question itself.

Many of the quotations in the Committee report attribute to former science advisers and to others and to a real frustration over their inability to do one of two things:

(1) Move an idea from the research phase to the development phase.

(2) Deploy scientific and engineering talent according to priority of national “problems." As examples of the former category, I would cite our attempts to launch a more vigorous attack on problems of ocean engineering, our attempts to develop operational weather modification, and our attempts to introduce innovations into the housing industry.

With regard to the first frustration, i.e., inability to go smoothly from research to production, I would like to point out that no one really knows how to manage this process efficiently. It is a continuing problem in industry. In providing funds, the backers of a project are often of the belief that all that is needed is to build a demonstration device and then go into production. If you want a specific example of this phenomenon, read the hearings fifteen or so years ago when the Office of Saline Water was established in the Interior Department. At that time, five demonstration plants were authorized. It took quite a few years for the engineers to swing that program around to a more reasonable balance of activity. The process of which they are most proud, reverse osmosis, wasn't even one of the five first chosen for "demonstration."

I have suggested before that the Congress should have independent sources of scientific and engineering advice. The National Academies of Science and of Engineering could be of considerable help in this connection and it has been encouraging to observe the effective relationship that has been established by your committee with the National Academy of Sciences. I believe these academies could be used more extensively in the evaluation of the processes of developmental engineering, and that the National Academy of Engineering could be used as advisors on development in the same way the National Academy of Sciences has been used as advisors in the initiation of projects.

As examples of the second kind of frustration, consider our inability to deploy our talents effectively in the fight against water pollution, air pollution, noise, flammable fabrics, and the decay of center city.

I suggest that the difficulties here are not managerial or even organizational. The difficulties arise because we have not until very recently, begun to make a commitment to these purposes. Whenever we have a clear-cut commitment to do something, we do find it possible to deploy our resources. Witness for example, our attack upon the conquest of space. Witness our attack on the problem of producing lower cost electricity from nuclear processes. These commitments and the necessary funding occurred without a prior centralization of the control. The prerequisite was that we had decided to do something.

There are many purposes to which the Government can put development. While industry uses development to make products which it can sell at a profit, the Government attempts to use R&D to assist in meeting social ends. We try to apply better methods to our urban problems, as in our attempts to improve housing and transportation. We deploy scientific talent to combat environmental decay, as in our studies of water pollution. We use research and development as the basis for economic improvement, as in our activities to develop better standards. We use research and development to provide for our common defense. We use research and development to extend or conserve our natural resources ; atomic energy is meant to extend our energy resources, water purification is meant to extend our water resources, waterway development is meant to extend resources for transportation. But in many cases the barriers to the application of this technology lie far beyond the question of the management of the development process. They involve the absence of a real commitment.

I stress these points because it seems to me that it is extremely difficult to discuss the control of development until we determine in advance, and are committed to, the purposes for which control will be effected.

It seems useful, therefore, to discuss Federal goals which might well be pursued through science and technology. It is useful to have several goals which may be pursued without necessarily stating which ones are first priority. Such an approach increases the support a particular project can have.

For example, it is often said that science ought to be regarded as a cultural contribution and supported simply because it is a means of strengthening the American culture. This is not an irrelevant view. Scientific research is one of the few international activities which may be discussed by men from entirely different cultures, with some hope of reaching an accord. In the end, I think that few are inclined to support science at a very strong level for this reason alone. Science and technology are supported as a means to accomplish improvements in human existence with the contributions to culture as a byproduct.

Another stated goal is the support of education. This is a valid and important objective. It costs almost as much to train a doctoral candidate in engineering, as it does an M.D. The schools cannot do it unaided. In the sciences and in engi

neering, support for graduate education has been available from NSF and from the AFOSR, ARPA, ONR, NIH, OSW, AEC and NASA, among others in connection with their support of academic research. I would urge that the NSF be strengthened. It has been especially effective in support of scientific education. It is time, however, for us to examine what the patterns of support for our schools and colleges has done and is doing to our educational system. While many leaders plead for engineers and scientists to turn their attention to urban problems, those who have had leadership responsibilities in the colleges have been hard pressed to find the funds to support these ventures. None of the agencies I have named above has a primary mission in the city. Only recently has HUD gotten into the field and with better funding of their mission, I would expect to see a beneficial impact on our schools and colleges. For one thing, our youngsters are demanding relevancy to contemporary social problems in their education ; this is a way to respond to their legitimate concerns.

The two goals, cultural and educational, which I have mentioned are important but do not justify Federal outlays of billions of dollars. The major objective is to enhance the well being of our people. Science and technology, through engineering, have enriched our lives beyond the wildest dreams of previous generations. They have also created problems beyond our wildest nightmares. I remind you that our problems are in large measure a result of our own success. We were successful in diminishing labor requirements on the farm—so our cities have grown like Topsy. We have automated our assembly lines and workers no longer are fulfilled by their labors. We have innovated at a dizzy pace, producing new goods and services, but this requires re-education of the men and women of our labor force or they are rendered unemployable.

Most significant research and development programs necessarily impinge upon the lives of citizens in ways which go beyond the competence of scientists and engineers to form valid judgments unaided by others in our society. Value judgments ought to be applied which are representative of the people at large. I believe that the consequences of a research and development program ought to be explored by representatives of the legislature, long before the operations are put into effect. Unfortunately, it takes time to communicate between scientists and engineers on one hand, and legislators on the other. One of the most significant efforts at communication occurred when the Atomic Energy Commission was first formed. You will recall that in those days teams of scientists came to Washington and talked earnestly with Representatives and Senators, teaching them the many new ideas which had to be mastered before legislation on atomic energy could be passed. But as I said, it takes time to communicate and the way we now carry out our legislative processes and decide upon our research and development funding, on an annual basis often does not afford sufficient time to discuss projects fully.

Mr. Chairman, in closing let me summarize the main arguments I have presented :

(1) Research, development and engineering require different managerial styles.

(2) Do not centralize research.
(3) Development needs good engineering management.

(4) The prerequisite to good management is a clear-cut statement of the goal.

(5) Make greater use of the National Academy of Engineering to advise on the development process.

(6) Seek better ways to involve the legislature in the goal setting phase of development projects.



Dr. TRIBUS. Mr. Chairman and members of the subcommittee, your committee report poses the question: "Should the science activities of the Federal Government be centralized ?" My answer to this question

will be as direct as the response of the young lady, who when asked, "Do you have trouble making up your mind ?" "replied, “Well, yes and no.”

The question of the control of science and engineering is too complex to admit a simple "yes" or "no" response. I agree on the importance of the questions raised in the committee report. Their very importance suggests that we ought to analyze carefully the alternatives actually available to us, the problems which prompted the questions in the first place, and how we might feel toward possible outcomes.

Mr. DADDARIO. Dr. Tribus, you will notice in our report that we took somewhat that same attitude. We really didn't ask anybody to say “Yes” or “No” to one specific set of facts, but rather set out a whole series of proposals diverse enough to include where we are, to create a central position where a czar might sit, as Mr. Waggonner has put it. So we find ourselves somewhat in the same position as you.

Dr. TRIBUS. C. P. Snow and others have called attention to a growing cultural dichotomy between science and the humanities. I feel this type of attitude is generated in part by the fear that science and technology might become ends in themselves. This fear must be disabused. Science and technology are not ends in themselves. The process which controls science and technology must be responsive to human needs, and I cannot emphasize this point too strongly.

The nature of control of our programs should be appropriate to the objectives. Our judgment of any type of control will be dependent upon our view of the role of the Federal Government in supporting research and development. Without agreement as to the responsibilities of Government, there will be no guidelines for the management research and development projects. Different expectations regarding the role of the Government will lead to different evaluations of the effectiveness of any arrangement, centralized or not.

I might insert here the remark that I have sat upon National Science Foundation committees that were concerned with giving funds to universities. And we were often puzzled as to whether we should give this money to a center of excellence where the best research would be done, or whether we should give this money as a social welfare gesture to help an institution which was poor in research. And the inability of those who were guiding us at the moment to tell us what our objective was made it extremely difficult for us to render good advice.

Now, in the private sector, research and development is usually carried out with a view toward a specific end product. The Federal Government, on the other hand, carries out research and development for a wide variety of reasons. The Federal Government conducts research to promulgate standards; for example, standards for auto safety, or for air pollution. The Federal Government conducts research and development in order to provide information for evaluation; for example, to guide purchasing by GSA. The Federal Government subsidizes the education of most of the engineers and scientists who study at the graduate level in the United States.

For example, NASA and the AEC research and development contracts with universities have also served as a means of enticing young men into new fields. I have often discussed contracting with representa

« PreviousContinue »