Recommendations II: Strengthening Physics Research—Partnerships, Federal Science Agencies, and Physics Information
Any human enterprise must adapt to new opportunities as well as new constraints if it is to stay effective. This is especially true of scientific research, where new opportunities emerge frequently from past accomplishment and technological change and where possibilities depend acutely on available resources. In physics, there was never a time when attention to these changes was more important than it is now. Accordingly, the committee has formulated four additional recommendations focused on the strength of the physics research enterprise. The first deals with the increasingly important role of partnerships among universities, industry, and national laboratories. The next two pertain to the stewardship of federal science agencies. The final recommendation pertains to the rapidly changing role of information technology in physics research and education.
Universities are at the core of the U.S. research system, combining the creation of scientific knowledge with its integration and dissemination. Long the envy of the world, they now face serious challenges. The cost of supporting research has grown rapidly in response to expanding scientific opportunities. To provide the faculty and the new infrastructure necessary for research, choices will have to be made and partnerships will have to be strengthened among universities and between universities and national laboratories.
National laboratories are a resource of enormous capability for the United States. They can cross disciplinary boundaries, serve scientists from universities and industry, and address problems of national importance. Realizing the full potential of this resource requires stable funding commit-
ments and effective cooperation with universities and industry. The national laboratories can play an important leadership role in this cooperation and in the formation of research partnerships.
For industry, partnerships with universities and the national laboratories provide important mutual benefits. Industry needs talented graduates to promote growth and innovation, and university faculty and students have much to gain from direct contact with the accelerating pace of technological change. For small companies that may not have R&D capabilities of their own, such partnerships can mean access to unique skills and facilities.
An urgent need is the development of workable intellectual property policies. Agreements must address inventions, patents, copyrights, and the restrictions of proprietary research. Difficulties naturally arise because of the different concerns of the participating institutions. Among large companies, cross-licensing arrangements are common to provide access to important patents. For small companies, exclusivity can make the difference between success and failure. Universities, often the sources of intellectual property but not the commercial users of it, can gain important income from its sale or licensing. National laboratories, which have concerns similar to those of universities, have benefited from legislation authorizing cooperative research and development agreements (CRADAs) for joint research projects with industries. In general, it has been difficult to achieve agreement on terms for research partnerships, particularly since the agreements are often negotiated by people not directly involved in the joint work.
A 1996 report of the Council on Competitiveness, Endless Frontier, Limited Resources: U.S. R&D Policy for Competitiveness, noted that “R&D partnerships hold the key to meeting the challenges that our nation now faces.” These partnerships include informal collaborations, exchange programs, facility sharing, and various formal relationships such as CRADAs. Technology is creating new modes for partnership, such as long-distance collaboration and information sharing. And partnerships are now more important than ever, since the nation's major industrial laboratories have been forced by divestiture and global economic forces to take a nearer-term, more focused approach to R&D.
Recommendation 6. The federal government, universities and their physics departments, and industry should encourage mutual interactions and partnerships, including industrial liaison programs with universities and national laboratories; visitor programs and adjunct faculty appointments in universities; and university and national laboratory internships and sabbaticals in industry. The
federal government should support these programs by helping to develop protocols for intellectual property issues in cooperative research.
FEDERAL SCIENCE AGENCIES
The base programs of the federal science agencies are structured around broad areas of research and are open to a wide range of proposals emerging from the scientific community. These proposals reflect the best ideas and ambitions of scientists throughout the country, and there is no better way for the agencies to ensure the strength and vitality of physics than to respond to the most creative of them. Some of the most exciting proposals are in emerging, interdisciplinary areas. Because the research activities supported by the base programs are diverse and because there is never a guarantee that an individual basic research effort will pay off, it has sometimes been difficult to make the case for the importance of these base programs. By contrast, special funding initiatives providing support for focused areas of research may be more easily understood and have recently been instrumental in securing new support for science agencies. Unless the balance between the base programs and special funding initiatives is monitored carefully, the diversity of high-quality physics research opportunities throughout the country will be jeopardized.
Recommendation 7. The federal government should assign a high priority to the broad support of core physics research, providing a healthy balance with special initiatives in focused research directions. Federal science agencies should continue to ensure a foundation that is diverse, evolving, and supportive of promising and creative research.
The federal government supports scientific research for the broad goal of general societal good and for specific programmatic goals. Predicting which researchers and research projects will best accomplish these goals can never be perfect. The peer review system relies on practicing scientists to advise on the feasibility of proposed research, the competence of the
researcher to carry out that work, and the potential impact of the work on science and technology. This system has served our nation well. Its success has convinced funding agencies in many other countries to use it as a model. Many have even gone so far as to make their funding decisions based on reviews of their own country's research programs by U.S. experts familiar with the peer review process in this country.
While the peer review system has worked well, federal science agencies and scientists must always be vigilant for possible abuses of the system. Care should be taken to ensure that reviews are based on scientific merit and are not influenced by other considerations, such as personal connections or prejudices. It is also the case that truly innovative work is often controversial and will naturally generate some mixed reviews. To avoid excessive conservatism, the peer review system must be flexible enough not to demand universally favorable reviews.
Recommendation 8. The peer review advisory process for the allocation of federal government support for scientific research has served our nation well over many decades and is a model worldwide for government investment in research. The peer review process should be maintained as the principal factor in determining how federal research funds are awarded.
Enabled by greatly increased capabilities in computation and data storage, physics questions are now being addressed by mining data from large databases. For example, in the next few years astrophysical data will be obtained and stored at a rate exceeding several thousand billion bytes per day. Similar data rates are expected from experiments at the new generation of high-energy colliders for nuclear and particle physics. The development of techniques for storing and accessing these enormous repositories of information is becoming essential for basic physics research. As the technical challenges are met, new modalities for data sharing and dissemination of research results become possible.
Concomitant with the rise in importance of large databases containing experimental results has been the increasing reliance on databases for the rapid dissemination of research results. For example, submissions to the archive of physics research based at Los Alamos National Laboratory increased by a factor of nearly 10 over the past decade. The rate of access to
this archive is equally impressive, having reached several hundred thousand per day. The archive has become an indispensable tool for conducting research in physics.
At the same time, the American Physical Society has dramatically improved its use of information technology in the publication of research results, establishing important links with the Los Alamos archive. The society is to be commended for reducing the costs of publishing its journals and for supporting the needs of the physics community. It should continue to do so, promoting the use of electronic methods and asserting its preeminent role in the processes of peer review and validation of research results.
Recommendation 9. The federal government, together with the physics community, should develop a coordinated approach for the support of bibliographic and experimental databases and data-mining tools. The use of open standards to foster mutual compatibility of all databases should be stressed. Physicists should be encouraged to make use of these information technology tools for education as well as research. The bibliographic archive based at Los Alamos National Laboratory has played an important role and it should continue to be supported.