The National Science Foundation: Serving the Nation in the 21st Century
An MIT Viewpoint for Consideration by the Special Commission on the Future of the National Science Foundation
The Mission of NSF
The mission of the National Science Foundation should be to ensure the long-term health of US science and engineering as an essential prerequisite for a vibrant national future. A central objective must be to secure and wisely disburse support for basic science and engineering research and education in pursuit of this mission. Fulfilling this mission will provide the nation with both new knowledge and a cadre of educated individuals prepared, through their independent careers, to address fundamental, industrial and societal challenges through science and technology.
NSF's modest budget of under $3B understates the importance of its role in the US R&D enterprise. It should be the agency with the broadest and longest view of science and technology in the national interest. Its first responsibility should be to see to the health of basic research, i.e., research that is inspired by intellectual opportunity and the quest for fundamental understanding, and not necessarily by predetermined practical objectives. This should be accomplished primarily by supporting research and graduate education. However, the environment, challenges and responsibilities of science and technology are changing, and this requires new mechanisms for accomplishing the Foundation's core objective and suggests new ramifications of its mission.
A Time of Change
The changing challenges of the post-cold war era are well known and present an opportunity to reexamine the nation's science and technology system. It should be emphasized that it is the entire system of government, industrial and academic components that is in need of attention, and we believe that the NSF review should be conducted in this context. A few of the many reasons for reexamination and potential change are:
1. The boundaries between basic and applied research have blurred and often disappeared.
2. The scale, complexity and interdisciplinary nature of many of the most interesting and challenging areas of frontier research are inconsistent with the currently dominant modes of research organization and funding.
3. "Economic security" is overtaking military security as a primary rationale for federal support of research.
4. Congress and the public expect a strong return on their investment in R&D, with a primary measure being a clear, positive effect on the nation's ability to compete or lead in the world marketplace.
5. We must learn to more rapidly and effectively build commercial successes from basic advances in science and engineering research.
6. Industrial laboratories are changing, with strong declines in basic research as industrial R&D organizations move away from centralized, corporate laboratories to use of "cross-functional" teams including researchers to address near-term problems and product development.
7. Attitudes in the US, particularly the "not invented here syndrome," often limit the effective use by development engineers of basic research information.
8. No single organization, indeed no single country, can now be completely dominant or self-contained in technological areas of economic significance.
9. The traditional linear path of basic research, development and finally commercialization no longer exists.
10. Shorter product life cycles and rapid reverse engineering discourage industrial investment in basic research.
11. Existing relationships among major research partners (government, industry and universities) are often perceived to involve excessive bureaucratic obstacles and impose excessive financial leveraging requirements on universities.
First and foremost, NSF must meet its central objective of supporting education and basic research in science and engineering. Whatever changes lie ahead, the NSF must rededicate itself to supporting the best ideas and people, wherever they are found, rather than supporting only bureaucratically determined areas and activities. Investigator-initiated projects will remain at the core of scientific creativity and discovery in the future as in the past. The disciplines must remain strong, but must not be allowed to "define away" exciting and important new areas of inquiry and activity that do not fit the traditional mold.
Once this central objective is secured, we recommend that the following be implemented:
1. NSF should develop better means to couple frontier research to technological development. Developing the attitudes and aptitudes needed to translate new knowledge from research to practical ends should be a part of the education of scientists and engineers at all levels. NSF should encourage strong partnerships with industry, but studies must first elucidate difficulties in technology transfer and establish realistic expectations of the practical benefits stemming from basic research. It should be noted that it often takes many years for new scientific knowledge to become technologically and commercially important.
2. NSF should create programs intended to fulfill industry's future personnel needs in science and engineering. In designing such programs NSF should foster the development of partnerships between universities and industries, to bring about common understanding of industry's needs and the university's role in fulfilling them. Both research and education need to be emphasized in connection with making effective linkages between US universities and industry. (We note as a model the MIT Leaders for Manufacturing Program, a true partnership of academia and US industry to educate a new class of manufacturing engineers and managers.) It is also crucial to this endeavor that explicit programs to further the participation of women, minorities and the handicapped in science and engineering be continued and improved.
3. NSF should organize to support programs with the parallelism and disciplinary diversity to address global-scale issues such as environmental change and telecommunications. Such programs will develop and build appreciation for teamwork.
4. NSF should establish ways of taking advantage of basic and applied knowledge generated elsewhere. Conscious efforts to overcome the "not invented here syndrome" should be included.
5. NSF should develop programs that support and encourage exchange of personnel between industry and academia to undertake cooperative research, both basic and applied. The individuals involved need to be encouraged to develop genuine partnerships, and it is these people-not the NSF-who should select and drive the research areas. There may be merit in considering having industrial development personnel spend time on campuses, in addition to frontier industrial researchers. Such partnerships should serve the needs of both large corporations and small start-up companies. [A prototype endeavor might be the DARPA-supported Consortium for Superconducting Electronics involving AT&T; IBM; and a small company, Conductus, in a cooperative effort with MIT (including Lincoln Laboratory), Boston University, Cornell and SUNY Stony Brook.]
We encourage continuing support of education by the NSF. Its primary role should be in graduate education through fellowships and research support, but it should also aggressively continue its efforts to enhance engineering and science education at the undergraduate level across the broad spectrum of colleges and universities. It should also play a substantial role of coordination and leadership in bringing about systemic change in science and mathematics education at the K-12 level.
The current environment will rightly encourage the NSF to explicitly consider ways to optimize technology transfer from universities to industry. We believe that the most effective mechanism for transfer of technology is the training of skilled and well-educated students. But there are additional ways of turning discovery into benefits to society, e.g. the development of new companies. Having observed many universities try to do this through the establishment of complicated centers, institutes and initiatives, MIT respectfully suggests that much is to be learned from the simpler and highly effective approach at MIT. The MIT Technology Licensing Office is currently making about 50 licensing agreements per year, has 350 active licenses, has assisted in the founding of about 40 new companies since 1987, and continues to assist the start-up of 6-10 new companies per year. A high density of high quality science and engineering research, long term support, and support in amounts consistent with the quality and promise of the research are all critical elements of success in this endeavor. We would be pleased to share our experience in detail with the Commission, the NSB or the NSF.
We are pleased that the National Science Foundation's mission for the future is being explored explicitly and carefully. Change is needed. But there is a possible vision of a future that must not be allowed to become reality: We must not find a new commission convened 10 or 20 years from now because our nation has learned to reduce product development cycles and manufacture high-quality products through continuous improvement, but is panicked because we have lost our edge in creativity and find that all the new ideas that are driving a knowledge-intensive world are generated elsewhere.
A version of this
article appeared in the
November 4, 1992
issue of MIT Tech Talk (Volume