2
Big Science, Little Science, and Their Relation to Space Physics
It has been the nature of science to grow—and to grow rapidly, outstripping population growth. De Solla Price [7] has shown that science has been characterized by an exponential growth rate for the past 300 years. This growth, measured by various manpower and publication parameters, is characterized by a doubling period of 10 to 20 years. Data analyzed by De Solla Price included scientific manpower, number of scientific periodicals, numbers of abstracts for various science fields, and citations. While the absolute values of these growth rates display a range of uncertainty, the general result is that the growth of science has been both long and rapid. This rapid growth is characteristic of all scientific subfields, old and new.
To appreciate how rapid a growth this is, note that the exponential growth of the general population shows a doubling period of about 50 years. Using 15 years to denote the doubling period for science, the ratio of the number of scientists to the general population doubles about every 20 years. Clearly, this trend cannot be sustained indefinitely. In fact, De Solla Price suggests that the problems facing science at this time are a reflection of its unusually long and rapid growth, a growth that, when compared with the much slower growth in the general population, may finally be straining the present economic fabric of society.
In this chapter the committee presents its views on the concepts and characteristics of big and little science as they pertain to the field of space physics. Each presents unique opportunities and challenges, and we conclude that both elements must be present for a research field to advance vigorously and productively.
CONCEPTS OF BIG AND LITTLE SCIENCE
In many discussions the concepts of big and little science are presented in near mythical terms—terms that cloud the complexity of the issues involved. Little science is usually represented by the lone researcher working in the laboratory on self-chosen problems, generally oblivious to the needs and/or requests of society. Big science, on the other hand, is often envisioned as a huge project or institute, managed by a bloated bureaucracy that directs, usually by committee, the scientific paths of many researchers. These are unsatisfactory and largely inaccurate generalizations that have led to more sterile argument than productive discussion.
One of the main reasons for this situation is that there is no absolute definition of big or little science. There seems to be a tendency in experimental science for small endeavors to evolve into large ones. Therefore, the bigness or smallness of any given scientific effort will depend on when it is observed within the evolution of its scientific subfield. In addition, the perceived size of a scientific project will vary from one subfield to another, as well as from one funding agency to another. What is considered a small satellite project is a very large project for rocketry or ballooning; what is a small project for the National Aeronautics and Space Administration (NASA) is generally a large project for the National Science Foundation. Furthermore, what is considered a small project today generally was thought to have been a large project years ago. This latter effect—the time dependence of the accepted measures of big and little project sizes—is a strong function of technological advances in the field. For example, today's desktop computers far outstrip the capabilities of the best mainframes of two decades ago—the big computer of yesterday is the little computer of today. A similar evolution has occurred in the space physics experimental arena, with the result that even today's small experiments are more sensitive, capable, complex, and expensive than those considered large in earlier years.
Although it is not possible to formulate accurate, universal definitions of big science and little science, it is possible to recognize each at a given point in time, in a particular subfield, and within a specific funding agency. The discussion in this report is based on researchers' perceptions of what constitutes big and little science, even though, as mentioned above, these terms vary by agency, subfield, and time.
CHARACTERISTICS OF BIG AND LITTLE SCIENCE
Big science and little science are characterized by very different needs, capabilities, and difficulties. In order that a proper balance between them be approximated in a given subfield, it is important to recognize how their respective strengths support the research objectives of the field.
Large projects are required for that unique class of science problems that
can be pursued only by using large, complex facilities and platforms, extensive campaigns, or multipoint observations. Small projects, typically pursued by many diverse investigators, are required for the steady progress and evolution of the field, as well as the unexpected results that often dramatically alter current perspectives. With an appropriate balance, there can be a strong synergism between large and small science that greatly enhances the productivity of the field.
Table 2.1 presents a concise list of some current characteristics of big and little science, as viewed from the space physics perspective. Because the terse
TABLE 2.1 Some Current Characteristics of Big and Little Space Physics Science
Big Science |
Little Science |
Broad set of goals |
Specific goal |
Interdisciplinary problems |
Discipline-oriented problems |
Scientific goals defined by committee |
Scientific goals defined by individual researcher/small group |
Researchers selected to fulfill program goals |
Researcher sets program goals |
Long implementation time |
Short implementation time |
Infrequent opportunities |
More frequent opportunities |
Large, complex management structure |
Minimal management structure |
High cost |
Relatively low cost |
Highly variable resource time line |
Relatively stable resource time line |
New-start funding process |
Base funding |
Supports project managers, engineers, administrators; science support comes at end of long planning, selling, implementation phases |
Supports science community throughout project |
Graduate student support data analysis phase |
Graduate student support through during entire project lifetime |
Dominant and increasing share of budget |
Minor and decreasing share of budget |
phrasing required for the table cannot convey the full complexity of these issues, an expanded discussion of each item is presented below.
Goals
Big science programs generally pursue broad sets of scientific goals that span the interests of several subfields. These goals are often backed by an influential constituency. Such programs are characterized by size (of both personnel and sheer physical infrastructure), complexity, and/or the numbers of experimental opportunities provided. Small science programs tend to address limited scientific goals, providing answers to specific science problems of importance in their research field.
Interdisciplinary Problems
Large programs represent an effective (and at times the only) way to pursue interdisciplinary science problems. Small science programs generally focus on problems within a single scientific discipline.
Project Definition
Because of the broad set of goals involved, definition of the science in large programs is accomplished through the use of committees representing all pertinent elements of the research field. In small science programs the more limited scientific goals are defined by the individual researcher and/or the small group involved.
Investigator Selection
Investigators in large programs are selected to fulfill the scientific goals set forth by the committee defining the program. In this sense big programs are often thought of as ''managed'' programs. However, it should be noted that a significant amount of independent research is often supported by such programs. In small programs, investigators are funded on the basis of the science that they propose within the program they have defined.
Implementation Time
Chapter 6 shows that along with the growth in program size a major increase in program implementation time has occurred. While small programs generally require shorter implementation times, Chapter 6 also shows that even they are experiencing implementation delays.
Frequency
Large programs necessarily occur infrequently because of their cost and long implementation times. Small programs can be supported at a much higher frequency.
Management
Large programs generally use large and complex management structures, while small programs are more often characterized by smaller, streamlined management structures. (Recently, large program management requirements have been increasingly applied to small programs with unfortunate results, as discussed in Chapter 6.)
Costs
Large programs are expensive; small programs are less expensive. However, it is important to remember that big science and little science vary from subfield to subfield, agency to agency, and notably in time. For example, the NASA Global Geospace Sciences mission, costing approximately $400 million, was thought of as a large mission when it was formally defined in 1988; today it is still considered a large mission in many quarters of space physics. However, a 1991 planning study [9] indicated that, at that time, NASA considered missions in the $300 million price range to be moderate, again showing the relative nature of big and little in time. Similar trends appear in other funding agencies.
Time Line for Resources
Because of the long implementation times and large costs involved, big programs result in a peak/valley resource time line, especially with respect to the science community supported by the mission. Small programs, because they are usually supported from base funding, can provide in the aggregate a relatively stable resource time line and are thus an important factor in maintaining the science infrastructure of the field.
Funding Process
Large programs generally require new-start funding approval by Congress on a program-by-program basis. This represents a large infusion of new resources into the field. Small programs are almost exclusively funded from an agency's base or core research funds. At times, however, large national initiatives, such as the Global Change Research Program, have included support for both large and small programs.
Community Support
Big science programs, during their planning and implementation phases, support an extensive management, administrative, and engineering infrastructure. A smaller operations effort is required during the mission's operational phase. Direct research support becomes available at the end of a long planning, selling, and implementation phase. Small science projects generally support the science community directly throughout the project's lifetime.
Educational Support
Due to their long planning and implementation times, large space physics programs generally are not appropriate training grounds for students (graduate and undergraduate) until these programs enter their operational and data analysis phases. At this time a substantial opportunity becomes available for data analysis and interpretation. (Data in Chapter 4 suggest that one result of this characteristic is an increasing average age for experimentalists and a decreasing average age for data analysts in space physics over the years.) Student support in small science programs is possible, and often required, throughout the planning, implementation, and data analysis phases. This provides excellent hands-on experience in experimental research and scientific program management, as well as in data analysis and interpretation.
Resource Share
Over the past decade, big science programs have come to command a dominant and increasing share of available funds (see Chapter 5). Conversely, small programs now represent a minor and decreasing share of the budget. Perhaps more importantly, the research efforts of small science are very vulnerable (sometimes to the point of extinction) to even small percentage cost overruns in big science projects.
Historical Interactions
A look at the history of the field of space physics shows that both large and small science projects have been used to advance the field to its present state of knowledge. First, as will be shown in Chapter 4, the field itself has grown over the past few decades from a small band of pioneers probing the mysteries of outer space with rockets and balloons to a community of several thousand researchers using sophisticated tools to study the space environment from the earth to the stars. As further indicated in Chapter 6, both small and large projects (as defined at the time) were used from the earliest days of what we now call space physics. Rockets and balloons played (and continue to play) a vital role in
studies of the upper atmosphere, ionosphere, aurora, and cosmic rays. These studies were greatly extended by the early satellite programs, judged large at the time. These satellite projects in turn evolved into big and little components, with the large platforms providing a capability to perform larger and more complex measurements than ever dreamed of before.
Space physics advanced rapidly during this period. Little science not only supported big science with the results of its research and discoveries but often itself evolved into big science operations. In a complementary fashion, big science supported little science by providing the platforms for experiments and data for many additional researchers and/or groups. A synergistic relationship existed whereby everyone seemed to benefit. Over the course of the 1980s, this symbiosis broke down. In large part, this report attempts to answer the question: "What went wrong?"
SUMMARY
Science grows, and in the past it has grown rapidly. Little science efforts often grow into larger ones, with the result that the recognition of what is big science and little science changes not only from field to field and agency to agency but also continuously with time. Current characteristics of big and little science, in the field of space physics, can be loosely defined. Big science projects attack broad problems with sophisticated technology and bring with them complex management structures, long implementation times, and high price tags. Small science projects involve individuals, or small teams of researchers, pursuing specific research goals via relatively inexpensive experiments that can be rapidly implemented.
We have seen that in the past big and little science projects in space physics were supportive of one another and synergistic in the research being pursued. Results of small science programs often motivated and formed the rationale for big programs; these big programs in turn provided many additional opportunities for the individual researcher.
Over the past decade tensions have developed between large and small science in the space physics community. Clearly, there can be no either/or; an appropriate dynamic balance between the two approaches must be found. This balance will vary both in time and from subfield to subfield. It is a balance that must be established and continually updated by the space physics research community and reinforced by the funding agencies.