Issues in Science and Technology, Online A Publication of National Academy of Sciences National Academy of Engineering Institute of Medicine University of Texas at Dallas
  • Home
  • Back Issues
  • Search


Rethinking What Research Government Should Fund

GERALD HOLTON

GERHARD SONNERT

A Vision of Jeffersonian Science

Jefferson's decision to fund the Lewis and Clark expedition reflected a philosophy of government support for science that makes good sense today.

The public attitude toward science is still largely positive in the United States; but for a vocal minority, the fear of risks and even catastrophes that might result from scientific progress has become paramount. Additionally, in what is called the "science wars," central claims of scientific epistemology have come under attack by nonscientists in the universities. Some portions of the political sector consider basic scientific research far less worthy of government support than applied research, whereas other politicians castigate the support of applied research as "corporate welfare."

Amid the choir of dissonant voices, Congress has shown interest in developing what is being called "a new contract between science and society" for the post­Cold War era. As the late Representative George E. Brown, Jr., stated, "A new science policy should articulate the public's interest in supporting science--the goals and values the public should expect of the scientific enterprise." Whatever the outcome, the way science has been supported during the past decades, the motivation for such support, and the priorities for spending are likely to undergo changes, with consequences that may well test the high standing that U.S. science has achieved over the past half century.

In this situation of widespread soul-searching, our aim is to propose an imperative for an invigorated science policy that adds to the well-established arguments for government-sponsored basic scientific research. In a novel way, that imperative tightly couples basic research with the national interest. The seemingly quite opposite two main types of science research projects that have been vying for support in the past and to this day are often called basic or "curiosity-driven" versus applied or "mission-oriented." Although these common characterizations have some usefulness, they harbor two crucial flaws. The first is that in actual practice these two contenders usually interact and collaborate closely, despite what the most fervent advocates of either type may think. The history of science clearly teaches that many of the great discoveries that ultimately turned out to have beneficial effects for society were motivated by pure curiosity with no thought given to such benefits; likewise, the history of technology recounts magnificent achievements in basic science by those who embarked on their work with practical or developmental interests.

As the scientist-statesman Harvey Brooks commented, we should really be talking about a "seamless web." The historian's eye perceives the seemingly unrelated pursuits of basic knowledge, technology, and instrument-oriented developments in today's practice of science to be a single, tightly-woven fabric. Harold Varmus, the director of the National Institutes of Health (NIH), eloquently acknowledged the close association of the more applied biomedical advances with progress in the more basic sciences: "Most of the revolutionary changes that have occurred in biology and medicine are rooted in new methods. Those, in turn, are usually rooted in fundamental discoveries in many different fields. Some of these are so obvious that we lose sight of them--like the role of nuclear physics in producing radioisotopes essential for most of modern medicine." Varmus went on to cite a host of other examples that outline the seamless web between medicine and a wide range of basic science disciplines.

The second important flaw in the usual antithesis is that these two widespread and ancient modes of thinking about science, pure versus applied, have tended to displace and derogate a third way that combines aspects of the two. This third mode now deserves the attention of researchers and policymakers. But we by no means advocate that the third mode replace the other two modes. Science policy should never withdraw from either basic or applied science. We argue that the addition of the third mode to an integrated framework of science policy would contribute tremendously to mobilizing widespread support for science and to propelling societal as well as scientific progress. Before we turn to a discussion of it, we will briefly survey the other two modes of scientific research.

Newtonian and Baconian research

The concept of pursuing scientific knowledge "for its own sake," letting oneself be guided chiefly by the sometimes overpowering inner necessity to follow one's curiosity, has been associated with the names of many of the greatest scientists, and most often with that of Isaac Newton. His Principia (1687) may well be said to have given the 17th-century Scientific Revolution its strongest forward thrust. It can be seen as the work of a scientist motivated by the abstract goal of eventually achieving complete intellectual "mastery of the world of sensations" (Max Planck's phrase). Newton's program has been identified with the search for omniscience concerning the world accessible to experience and experiment, and hence with the primary aim of developing a scientific world picture within which all parts of science cohere. In other words, it is motivated by a desire for better and more comprehensive scientific knowledge. That approach to science can be called the Newtonian mode. In this mode, the hope for practical and benign applications of the knowledge gained in this way is a real but secondary consideration.

Turning now to the second of the main styles of scientific research, popularly identified as "mission-oriented," "applied," or "problem-solving," we find ourselves among those who might be said to follow the call of Francis Bacon, who urged the use of science not only for "knowledge of causes and secret motion of things," but also in the service of omnipotence: "the enlarging of the bounds of human empire, to the effecting of all things possible."

Research in the Baconian mode has been carried out more commonly in the laboratories of industry than of academe. Unlike basic research, mission-oriented research by definition hopes for practical, and preferably rapid, benefits; and it proceeds, where it can, by using existing knowledge to produce applications.

Jeffersonian research

Recognition of the third mode may open a new window of opportunity in the current reconsiderations, not least in Congress and the federal agencies, of what kinds of science are worth supporting. It is a conscious combination of aspects of the Newtonian and Baconian modes, and it is best characterized by the following formulation: The specific research project is motivated by placing it in an area of basic scientific ignorance that seems to lie at the heart of a social problem. The main goal is to remove that basic ignorance in an uncharted area of science and thereby to attain knowledge that will have a fair probability--even if it is years distant--of being brought to bear on a persistent, debilitating national (or international) problem.

An early and impressive example of this type of research was Thomas Jefferson's decision to launch the Lewis and Clark expedition into the western parts of the North American continent. Jefferson, who declared himself most happy when engaged in some scientific pursuit, understood that the expedition would serve basic science by bringing back maps and samples of unknown fauna and flora, as well as observations of the native inhabitants of that blank area on the map. At the same time, however, Jefferson realized that such knowledge would eventually be desperately needed for such practical purposes as establishing relations with the indigenous peoples and would further the eventual westward expansion of the burgeoning U.S. population. The expedition thus implied a dual-purpose style of research: basic scientific study of the best sort (suitable for an academic Ph.D. thesis, in modern terms) with no sure short-term payoff but targeted in an area where there was a recognized problem affecting society. We therefore call this style of research the Jeffersonian mode.

Congress has shown interest in developing what is being called "a new contract between science and society."

This third mode of research can provide a way to avoid the dichotomy of Newtonian versus Baconian styles of research, while supplementing both. In the process, it can make public support of all types of research more palatable to policymakers and taxpayers alike. It is, after all, not too hard to imagine basic research projects that hold the key to alleviating well-known societal dysfunctions. Even the "purest" scientist is likely to agree that much remains to be done in cognitive psychology; the biophysics and biochemistry involved in the process of conception; the neurophysiology of the senses such as hearing and sight; molecular transport across membranes; or the physics of nanodimensional structures, to name a few. The results of such basic work, one could plausibly expect, will give us in time a better grasp of complex social tasks such as, respectively, childhood education, family planning, improving the quality of life for handicapped people, the design of food plants that can use brackish water, and improved communication devices.

Other research areas suited to the Jeffersonian mode would include the physical chemistry of the stratosphere; the complex and interdisciplinary study of global changes in climate and in biological diversity; that part of the theory of solid state that makes the more efficient working of photovoltaic cells still a puzzle; bacterial nitrogen fixation and the search for symbionts that might work with plants other than legumes; the mathematics of risk calculation for complex structures; the physiological processes governing the aging cell; the sociology underlying the anxiety of some parts of the population about mathematics, technology, and science itself; or the anthropology and psychology of ancient tribal behavior that appears to persist to this day and may be at the base of genocide, racism, and war in our time.

It is of course true that Jeffersonian arguments are already being made from time to time and from case to case, as problems of practical importance are used to justify federal support of basic science. For instance, current National Science Foundation­sponsored research in atmospheric chemistry and climate modeling is linked to the issue of global warming, and Department of Energy support for plasma science is justified as providing the basis for controlled fusion. NIH has been particularly successful in supporting Jeffersonian efforts in the area of health-related basic research. Yet what seems to be missing are an overarching theoretical rationale and institutional legitimization of Jeffersonian science within the federal research structure.

The current interest in rethinking science and technology policy beyond the confining dichotomy of basic versus applied research has spawned some efforts kindred to ours. In Donald Stokes's framework, the linkage of basic research and the national interest appeared in what he called "Pasteur's Quadrant," which overlaps to a degree with what we have termed the Jeffersonian mode. Our approach also heeds Lewis Branscomb's warning that the level of importance that utility considerations have in motivating research does not automatically determine the nature and fundamentality of the research carried out. Branscomb appropriately distinguishes two somewhat independent dimensions of how and why: the character of the research process itself (ranging from basic to problem-solving) and the motivation of the research sponsor (ranging from knowledge-seeking to concrete benefits). For instance, a basic research process, which for Branscomb comprises "intensely intellectual and creative activities with uncertain outcomes and risks, performed in laboratories where the researchers have a lot of freedom to explore and learn," may characterize research projects with no specific expectations of any practical applications, as well as projects that are clearly intended for application. Branscomb's category of research that is both motivated by practical needs and conducted as basic research is very similar to our concept of Jeffersonian science.

The Carter/Press initiative

Jeffersonian science is not an empty dream. A general survey of related science policy initiatives can be found in the article by Branscomb that follows this one. Here we briefly turn to a concrete 20th-century example of the attempt to institute a Jeffersonian research program on a large scale. Long neglected, that effort is eminently worth remembering as the covenant between science and society is being reevaluated.

Jeffersonian science can make public support of all types of research more palatable to policymakers and taxpayers alike.

In November 1977, at President Carter's request, Frank Press, presidential science adviser and director of the Office of Science and Technology Policy, polled the federal agencies about basic research questions whose solutions, in the view of these agencies, were expected to help the federal government significantly in fulfilling its mission. The resulting master list, which was assembled in early 1978, turned out to be a remarkable collection of about 80 research questions that the heads of the participating federal government agencies (including the Departments of Agriculture, Defense, Energy, and State and the National Aeronautics and Space Administration) at that time considered good science (good, here, in the sense of expected eventual practical pay-offs) but which, at the same time, would resonate with the intrinsic standards of good basic science within the scientific community. It should be added here that the agency heads could make meaningful scientific suggestions thanks in good part to two of Press's predecessors, science advisers Jerome Wiesner and George Kistiakowsky. They had helped to build serious science research capacities into the various federal mission agencies, thus ensuring that highly competent advice was available from staff scientists within the agencies.

Consider, for instance, this question from the Department of Agriculture: "What are mechanisms within body cells which provide immunity to disease? Research on how cell-mediated immunity strengthens and relates to other known mechanisms is needed to more adequately protect humans and animals from disease." That question, framed in 1978 as a basic research question, was to become a life-and-death issue for millions only a few years later with the onset of the AIDS epidemic. This selection of a research topic illustrates that Press's Jeffersonian initiative was able in advance to target a basic research issue whose potential benefits were understood in principle at the time but whose dramatic magnitude could not have been foreseen (and might well not have been targeted in a narrow application-oriented research program).

Other remarkable basic research questions included one by the Department of Energy about the effects of atmospheric carbon dioxide concentrations on the climate and on global social, economic, and political structures, as well as one by the Department of Defense about superconductivity at higher temperatures, almost a decade before the sensational breakthrough in this area.

A Jeffersonian revival

The Carter-Press initiative quickly slid into oblivion when Carter was not elected to a second term, yet it should not be forgotten. A revitalization of the Jeffersonian mode of science would provide a promising additional model for future science policies, one that would be especially relevant in the current state of disorientation about the role of science in society.

For many scientists, a Jeffersonian agenda would be liberating. Scientists who intended to do basic research in the defined areas of national interest would be shielded from pressures to demonstrate the social usefulness of their specific projects in their grant applications. Once these areas of interest were determined, the awards of research grants could proceed according to strictly "science-internal" standards of merit.

Moreover, a Jeffersonian agenda provides an overarching rationale for the government support of basic research that is both theoretically sound and can be easily understood by the public. It defuses the increasingly heard charge that science is not sufficiently concerned with "useful" applications, for this third mode of research is precisely located in the area where the national and international welfare is a main concern. The way basic research in the interest of health is already legitimized and supported under the auspices of NIH may well serve as a successful example that other sciences could adopt.

Finally, the strengthened public support for science induced by a visible and explicit Jeffersonian agenda is likely to generalize and transfer to other sectors of federal science policy. (Again, we do not advocate the total replacement of the Newtonian and Baconian modes by the Jeffersonian mode; all must be part of an integrated federal science policy.) Even abstract-minded high-energy physicists have learned the hard way that their funding depends on a generally favorable public attitude toward science as a whole. Moreover, they too can be proud of the use of the campus cyclotron for the production of radioisotopes for cancer treatment and the use of nuclear magnetic resonance or synchrotrons for imaging. Nor should we forget the current valued participation of pure theorists on the President's Science Advisory Committee and other important government panels; nor their sudden usefulness, with historic consequences, during World War I and World War II. From every perspective, ranging from the cultural role of science to national preparedness, even the "purest" scientists can continue to claim their share of the total support given to basic science. But that total can more easily be enlarged by the change we advocate in the public perception of what basic research can do for the needs of humankind.

Recommended reading

L. M. Branscomb, "From Science Policy to Research Policy," in Investing in Innovation: Creating a Research and Innovation Policy That Works, ed. L. M. Branscomb and J. H. Keller (Cambridge, Mass.: MIT Press, 1990), 112­139.

H. Brooks, "The Changing Structure of the U.S. Research System, in H. Brooks and R. Schmitt, Current Science and Technology Issues: Two Perspectives, Occasional Paper No. 1 (Washington, D.C.: Graduate Program in Science, Technology, and Public Policy, George Washington University, 1985), 17­72.

G. Holton, The Advancement of Science, and Its Burdens (New York: Cambridge University Press, 1986) (second edition: Cambridge, Mass.: Harvard University Press).

G. Holton, "What Kinds of Science Are Worth Supporting? A New Look, and a New Mode" in The Great Ideas Today (Chicago, Ill.: Encyclopaedia Britannica, 1998), 106­136.

G. Holton, H. Chang, and E. Jurkowitz, "How a Scientific Discovery is Made: A Case History," American Scientist 84 (1996): 364­375.

House Committee on Science, Unlocking Our Future: Toward a New National Science Policy (the "Ehlers Report") (Washington, D.C.: September 1998).

D. E. Stokes, Pasteur's Quadrant: Basic Science and Technological Innovation (Washington, D.C.: Brookings Institution, 19XX).


Gerald Holton is Mallinckrodt Professor of Physics and professor of history of science (emeritus) and Gerhard Sonnert is research associate in physics at Harvard University.