Science Advocacy is Inevitable: Deal with It*

Margaret A. Shannon

Errol E. Meidinger

Roger N. Clark+

 


* Paper prepared for the annual meetings of the Society of American Foresters, November 11, 1996, Albuquerque, NM.

+ Margaret Shannon is Associate Professor of Public Administration, Syracuse University, Syracuse, NY 13244 (mshannon@acsu.buffalo.edu). Errol Meidinger is Professor of Law and Adjunct Professor of Sociology, State University of New York, Buffalo, NY 14260 (eemeid@acsu.buffalo.edu). Roger Clark is Program Manager for the People and Natural Resources Research and Development Program, PNW Research Station, U.S.D.A. Forest Service, Seattle, Washington 98105.


 

I. INTRODUCTION

The apparent advocacy activities of scientists have been a major topic of discussion in natural resource policy circles in recent years. Largely taken-for-granted in that discussion have been the assumptions that advocacy by scientists is inappropriate and that ways must be found to eliminate it, probably by more sharply defining and separating the roles of scientists, managers, and policy makers. This paper seeks to directly challenge those assumptions, first by clarifying the management and policy context in which the work of scientists has become increasingly controversial, then by detailing the ways in which scientific work is and always has been a form of advocacy. The concluding section of the paper argues that the appropriate response to the inevitability of scientific advocacy requires scientists, managers, and society to negotiate a changed set of expectations for science more in accord with the reality of modern scientific practice.

Much of the underlying work for this paper grows out of recent efforts by the Forest Service to address the increasingly contentious relationship of science activities to its policy and management activities. In November of 1995, a conference on bioregional assessments was held in Portland, Oregon called "At the Crossroads of Science, Management and Policy." Five case studies were prepared and presented by individuals who participated in them, and critiques were presented by scientists, managers and policy makers. For this conference, Shannon co-authored the case study of the Forest Ecosystem Management Assessment Report (FEMAT). She also presented a conceptual framework for considering alternative models of science in the policy process.

Meidinger conducted an extensive review of the published literature on science and policy (e.g., Anytpas and Meidinger 1996) as a part of a cooperative research project with the Pacific Northwest Research Station. This project examined the role of scientists in the policy processes by undertaking an in-depth study of the scientists involved in reshaping federal forest policythe policy processes in the Pacific Northwest.

Stemming from increasing internal concern with the roles of both scientists and managers in these science-policy assessments, the Forest Service undertook to develop guidelines for how to manage them. As part of the background information assembled for this effort, the authors conducted a series of ten case studies of recent examples of science intensive policy processes. Four cases were studied in depth: the Columbia River Basin Ecosystem Assessment (CRB), the Anadromous Fish Habitat Assessment (Pacfish), the Instream Flow Adjudication Assessment Process in Colorado (IF), and the Red Cockaded Woodpecker (RCW) policy process in the southeast. Six others received less intensive review: the Forest Ecosystem Management Assessment (FEMAT), the Interregional Habitat Conservation Process (IRH), the National Acid Precipitation Assessment Project (NAPAP), the Northern Goshawk Habitat Assessment (NG), the Southern Appalachian Ecosystem Assessment (SA), and the Tongass Land Management Planning Process (TLMP). The initial results of this research were reported and discussed at an intra-agency Science and Policy Seminar in Rennsalaerville, New York, on June 4-6, 1996, and are part of ongoing agency deliberations on the possibility of developing guidelines for the use of science in management and policy.

Although the authors have been involved in Forest Service research efforts, however, the positions expressed in this paper are not to be taken as expressions of agency analysis or policy; indeed they are likely to be viewed as quite controversial by many scientists, managers, policy makers, and citizens. Rather, this paper should be taken as one contribution to ongoing discussions which the authors believe are crucial to understanding and going beyond current conflicts.

 

 

II. THE CHANGING MANAGEMENT AND POLICY CONTEXT OF NATURAL RESOURCES SCIENCE

A. "The Fusion of Management and Science Will Enhance Accountability"

1. "Credible Scientific Information" -- The First Mission of the Division of Forestry (1876-1898)

In 1876, with an appropriation of $2000, Congress responded to growing societal concerns regarding deforestation and its effect on flood prevention, timber supply, recreation and aesthetics by authorizing Dr. Franklin Hough to undertake a national study of "annual amount of consumption, importation, and exportation of timber and other forest products, the probable supply for future wants, the best means adapted to their preservation and renewal, the influence of forests upon climate, and measures ... for the preservation and restoration or planting of forests ..." (Robinson 1975:2). This study provided the initial scientific basis for federal forest policy. Providing credible scientific information remained the primary mission even after the Bureau of Forestry was elevated to a Division of Forestry within the Department of Agriculture and Bernard Fernow was named as its head (Schiff 1962:1). Numerous research efforts during this period contributed to the utilization and conservation of forest resources. One familiar example is the work on how to preserve railroad ties which had to be replaced every few years. The search for and discovery of creosote as a preservative greatly extended the resistance of ties to rot (Peffer 1951).

2. "Forests For the Homebuilder First of All" -- The United States Forest Service is Established

Gifford Pinchot was named as head of the Division of Forestry in 1898, one year after Congress passed what is now called the "Organic Act" which defined and limited the purposes, management and uses of the forest reserves. Pinchot believed that for forestry to ensure a stable and continuous supply of timber, the forest reserves must be placed under forestry management, for the private timber companies had neither the interest or capacity to consider the long term. Thus, from the outset, his primary objective was to gain management control of the forest reserves and he believed that active management was the way to gain this control. Transfer of these lands occurred in 1905, and the Division was renamed the Forest Service in 1907.

In shifting the orientation of the new Forest Service from provider of scientific information to scientific forest management, Pinchot was in step with the Progressive Movement of the era (Hays 1959). The Progressive Movement embraced the idea of science and scientific management as the means to social and economic progress and security. However, the Progressives were men of action who zealously resisted societal forces for moving slowly and preserving traditions. They embraced the "gospel of efficiency" as the means to a better society. One must recall that this period from 1890 to 1920 was a tumultuous period in American history and that science was employed as a rational basis for changing traditional roles and relationships (Wiebe 1967)

Pinchot appreciated that the American Association for the Advancement of Science had first raised concerns over the nation’s timber supply. He rightfully credited a National Academy of Science report as originating federal forest policy (Ise 1920). Yet, while Pinchot was Chief, within the agency priorities the advancement of knowledge through research as a on-going part of management was sorely neglected, and indeed grew to be disdained, as an impediment to efficient administration. This legacy continues today in that managers once inspired by science to develop new management approaches, often make little provision for continued scientific effort to monitor outcomes and identify problems resulting from the application of management tools and techniques.

3. Science and Management: Integrate or Separate?

During this time, the role of the federal government as provider and user of scientific information grew expansively. Numerous federally funded research organizations were created. The Progressives used science as the means to creating and legitimating their agenda of progress. In forestry, this legacy is especially evident in the mission of the land grant universities and their mission to improve agriculture, home economics, fisheries, and forestry with scientific methods.

Forestry has always been both an arena of scientific investigation and a method for how forests should be managed. The history of forestry education clearly shows the tension between these two roles (cite - there is a dissertation just completed on this). Given the historic centrality of forests to national economic production, many nations supported forestry research organizations. The International Union of Forestry Research Organizations (IUFRO) brought scientists together from these organizations to share their discoveries. Because only forestry research organizations could be members of IUFRO, the United States had difficulty participating. As a result, there was broad interest in creating a forestry research organization in U.S. so as to participate in this international arena.

All of these considerations were present, when in 1928 Congress passed the McSweeney-McNary Act creating the third major branch of the Forest Service -- Research and Experimental Forestry. The Research Branch rapidly grew into a large and thriving branch of the agency. So central did research become in this period, that later when President Franklin D. Roosevelt tried to transfer the Forest Service to the department of Interior in 1940, even Pinchot protested saying, "The Forest Service is a research as well as an executive organization. This union of research and administration underlies the progress of forestry and the morale and efficiency of the Service. To separate the two would ruin the Service" (Schiff 1962:2-3). Pinchot also argued that the administrative integration of scientists meant that the scientists could not hide in ivory towers and disclaim responsibility for management consequences. Further, when science was directly utilized in developing management actions, scientists were less likely to permit data to accumulate without analysis and interpretation in an attempt to remain above controversy (Schiff 1962:3).

So, in 1928 with these salutary arguments in mind, many legislators accepted the integrationists’ arguments out of their own belief that the fusion of science and management would strengthen administrative accountability. This would occur, they felt, because the "uncompromising concern with honesty and integrity, the relentless search for truth, and the necessary freedom of inquiry would prevent managers from getting ‘into ruts’ and succumbing to the arrogant and intolerant attitudes characteristic of crusading organizations" (Schiff 1962:3). Integrationists also noted that there were many other sources of scientific investigation including other agencies as well as universities which would temper any co-optation of the scientists by the goals of management. In 1962, Ashley Schiff published his path-breaking study of the co-optation of fire research to support fire exclusion policies and water research certify the need for forests to control floods and preserve water flows in, Fire and Water: Scientific Heresy in the Forest Service. In it, he notes (p.4) that when deciding whether to keep research a part of the Forest Service, Congress believed that with an internal research branch any tendency to rely on policies contravening scientific principles would ultimately prove unsatisfactory, in not embarrassing, to the Service. This self-correcting mechanism would, thereby strengthen the willingness of administrators to respond to scientific information.

B. Managing Research: The Co-optation of Science to Policy

It is sobering that only eight years later, in 1935, on the occasion of his promotion to Associate Chief, then Assistant Chief of Research Earle Clapp wrote to the Chief F. A. Silcox, "I am leaving Research with an uneasy feeling about its future in the Forest Service which I know is shared by others. As yet is has not in my judgment been conclusively shown that it is possible, in a bureau such as the Forest Service, primarily administrative in its functions, to develop and permanently maintain a strong, effective research organization." Dr. Clapp then stated that he could leave his position assured that a strong Research Branch would be maintained if Silcox would approve several principles, including the centrality of research to the purpose of the agency. He then wrote that the reasons for a strong research branch included: providing the basic biological, social, economic, and other technical information necessary for the rapid and well-rounded progress of the whole forestry movement, to help make the Forest Service a technical organization which he believed it was not, and most importantly, to have "at all times in the Forest Service a group not under administrative domination, idealistic from the very nature of its work, ready when occasion demands to supply the criticism which the Forest Service needs to keep it alive and forward-looking, and also to perform the same function for American forestry as a whole" (Schiff 1962:13).

"... the time is already long overdue, major Forest Service policies and major national forestry policies will have to be based on fact and not on opinion, and this means basing them on research results. ... the Forest Service in its councils should recognize the men who collect and interpret these facts on a basis of equality with the men whose job it is to apply them. These recommendations are, I fully realize, pure heresy from the traditional administrative standpoint of the Forest Service" (Schiff 1962:14).

Schiff’s study of the management co-optation of fire and water research reminds us of the difficulties of maintaining the internal capacity for critique and change within an administrative organization. He finds, as did Philip Selznick in his dissertation research on the Tennessee Valley Authority published as The TVA and the Grassroots (1949), that the mechanisms of co-optation are simple, and difficult to overcome (Schiff 1962:Ch.5).

A brief list of a few of these mechanisms is provocative for, as we show in the next section, they remain the focus of concern today. Current concern with many of these same issues led to agency and Congressional examination of the relationships between scientists and managers, especially within the context of science-policy efforts resulting in dramatic shifts in federal land policy and management. The findings and conclusions from the case studies and workshop discussions echo the following conclusions drawn by Ashley Schiff in 1962.

When considering all of these mechanisms, it is easy wonder if change is needed. Is the solution the separation of the scientific enterprise from the administrative one? Studies of administration agree that no panacea exists. Rather, vigilant attention to maintaining an open and critical relationship is essential.

The mechanisms of co-optation listed above are not solved with a simple separation of research from administration. These mechanisms are at play elsewhere. For example, we see them present in the efforts of scientific disciplines to maintain their identity by constraining what problems are addressed by what methods, in the efforts of academic departments to maintain their budgetary influence by distinguishing their work from that of others within the university, and in the inability of researchers to embrace complex, multidisciplinary problems in an environment of fragmentation of discipline and funding. Thus, the problem is not carefully separating science from politics and the shoals of advocacy, but rather in recognizing that science is a complex enterprise. The simple view that scientists give policy makers a "bucket of facts" and quickly exit from the room is silly. Scientists are essential actors in shaping our understanding of the nature of problems, the way the world works, and the realms of potential solutions. They are necessarily political actors. The question is what kind?

 

C. Science-Policy Assessments: New Roles for Scientists and Managers

Current interest in the large scale scientific assessments in natural resources is the most recent episode in this historic debate. The central question remains: If the role of scientists is to leverage policy and management changes based on the authority of credible scientific research, what norms should govern them as they play this role?

In order to understand why these science-policy processes are demanding new roles from both scientists and managers, we have developed two simple typologies. The first typology explicates the nature of the problem as defined by managers, the second one does the same but from the point of view of the scientists. By way of background, it is useful to note that these typologies were conceptualized based on a series of questions posed by a group of researchers and managers within the Forest Service.

1. New Roles for Line Officers, Managers, and Technical Specialists

The typology in Figure 1 has two basic dimensions related to making and implementing decisions on the basis of information, including scientific information about biological, social and economic aspects. Does a manager have sufficient authority to make and carry out a decision, or is coordination with others necessary? Routine is the application of a given management prescription to a pre-defined problem. Management prescriptions are the result of routines and the ability to carry them out. This simplest case of management decision-making is often invoked as the preferred situation. Indeed, most models of managerial decision making view this as the ideal. The problem is that it almost never applies in real practice.

The concept of an unambiguous management prescription is an ideal, but the choice of definition of the problem is where managers constrain what prescriptions are applied. For example, timber sales nearly always receive a "no significant impact" evaluation in the environmental assessment stage on the basis that they are routine agency actions. The question of whether to harvest the trees is thereby avoided. The explicit role of agency managers is to follow routines and prescribe defined management actions. Herbert Kaufman’s dissertation research on administrative behavior, published as The Forest Ranger in 1960, termed this relationship "predetermined decisions."

 

 

One useful application of this typology is to enable us to see the dynamics of the organization. So, for example, when managers are confronted by surprise -- new problems, new information, new policy directives, and so on -- they embrace a process of adaptive management in order to search out new problem definitions and new solutions. In this search process, they work with scientists to experiment with potential solutions. It is worth noting that while beyond our focus on scientists in this paper, this is where democratic processes and representative bureaucracy are also important factors.

As resource management problems expand beyond single agencies much less single administrative units, managers must engage in some forms of coordination. Because of overlapping jurisdictions and intermingled land ownership, complex problems nearly always involve other agencies as well as tribes, state and local governments, land owners, communities, non-government organizations, and citizens. In situations where routine is possible because the problem is clearly defined and the authority of the various actors clear, various forms of interagency coordination are possible. Advisory councils and interagency task forces are common mechanisms for this situation.

It is when authority is unclear or unstated and the problem is ill-defined and complex that scientists and managers create new arrangements for investigating the nature of the problem and developing new policies. The science-policy assessments are examples of these new arrangements. Yet, it is important to note that while these assessments are becoming a part of the normal business of natural resource management, nonetheless the organizational imperative is to translate new policies into routine solutions, or management prescriptions.

The role problem for managers is different for line officers and technical specialists. Line officers must carry out the policies of the agency. What is their role in working with scientists to create new policies? Role conflict was evident in every interview we did with line officers; knowing when to help craft new solutions versus when to carry out routines until instructed differently presented a real dilemma. For management specialists the issue is different. When should they participate as scientists open to wherever inquiry takes them, and when must they carry out the management solutions routinely prescribed? Every interview and workshop discussion raised this issue. For administrators, the problem was the "management of specialists" who were viewed as having more loyalty to their specialty than to the agency. For specialists, it was a problem of being recognized and valued as scientists with the authority to bring new information to bear in developing new management prescriptions.

2. New Roles for Researchers

When research is directly related to informing management and policy decisions, several roles emerge depending on whether the problem fits within a single discipline and to what extent current knowledge is sufficient to inform scientific judgment. The typology shows that when existing knowledge is sufficient, then the role of scientists is one of expert judgment or "technology transfer" -- make the information accessible to managers. When the problem moves beyond existing knowledge but remains within the realm of traditional disciplinary science, then scientists undertake traditional research to develop new knowledge.

However, many resource management problems today involve multiple resources and interdisciplinary research efforts. When the state of the knowledge in the various disciplines is sufficient once combined with others, then scientists join in interdisciplinary teams to make expert judgments. It is when problems extend beyond traditional scientific research interests and require the definition of new problems for scientific investigation that scientists engage in interdisciplinary research and collaboration with managers and others in understanding the implications of different ways of defining the problem and different options for solving it.

 

 

 

 

These roles -- defining the problem and assessing options for solving it -- go beyond traditional images of research and open scientists to the charge of advocacy. Advocacy in this sense means that scientists pursue research which will support their preferred definition of the problem and solution to it. Thus, in the Forest Ecosystem Management Assessment Team scientists were charged with having chosen the options of a reserve approach to the problem of managing late successional forests because they didn’t trust managers not to employ routine methods of harvesting given the option to do so. As a result, attempts were made to discredit the scientific assessment on the basis of bias and advocacy. In every case we studied similar attempts to discredit the scientists working on these science-policy assessments were made by charging ‘advocacy.’

If the norms of science are open and unfettered inquiry, how do scientists avoid the charges of advocacy when they participate in assessment where current policy is no longer adequate and new policies are necessary? This question assumes a clear and unambiguous line between ‘pure science’ and ‘pure advocacy.’ In reality the distinction is fuzzy. Redefining the concept of old forests from decadent and useless to vital, indispensable components of healthy ecosystems discredited the management prescriptions based on the earlier view. Once discredited new policies were needed and the scientists who shaped the new conception were instrumental in defining new policies. This role is common to all of the assessments we studied and it is the origin of the charges of ‘advocacy science.’

 

 

III. ADVOCACY IN SCIENCE

The privileged role of science in modern society is often justified by its ostensible adherence to rigorous methods of data gathering and analysis and strict logical standards for assessing the validity and generalizability of findings. These methods and standards are seen to be implemented in a larger community of open communication which allocates rewards and punishments based on how well they are achieved (e.g., Popper 1961, Merton 1973, Fuchs 1993). For this reason, many scientists, policy makers, and observers view situations involving values, interests, or policy as anathema to science. Regardless of the overall accuracy of the "positivist" model of science, however, there is a large body of research indicating that values, interests, and advocacy often go hand-in-glove with science, and may be preconditions to its success in some cases. This section briefly summarizes key elements of that research in the form of a series of propositions supporting the conclusion that it is untenable to simply to isolate science from these factors.

 

A. The Social Value of Science.

1. Historically, science has developed and been supported because of its ability to achieve human purposes in the world. At the dawn of modern science Francis Bacon observed that the primary reason humans seek knowledge of nature is to "extend their dominion over things" and also argued that this goal would lead to the eventual triumph of science as the queen of knowledge (Bacon 1620). Soon thereafter, Rene Descartes, as close to a patron saint of modern science as there is, wrote about science:

"We can have useful knowledge by which, cognizant of the forces and actions of fire, water, air, the stars, the heavens, and all the other bodies that surround us, knowing them as distinctly as we know the various crafts of the artisans, we may be able to apply them in the same fashion to every use to which they’re suited, and thus make ourselves masters and possessors of Nature." (Descartes 1637).

2. Scientists frequently advocate the social importance of specific kinds of research. Common examples are the competition for limited research funding between disciplines such as silviculture and forest ecology, or the biological and social sciences, as well as within disciplines. This is partly, but not entirely attributable to the need to obtain research funding and as well as the desire to obtain credit for doing important research. In each case, scientists assert that their proposed research will do more to solve important problems than competing research. Of course, doing so requires a definition of important problems, and this requires scientists to advocate normative positions.

3. Different kinds of knowledge are useful for different purposes. Therefore a choice to obtain one kind of knowledge as opposed to a different kind of knowledge is in effect a choice to prefer one kind of purpose over another. As a result of research support choices, data to support particular kinds of policies are often produced, whereas data that might support contrary policies are not produced. For example, if the only research available had been on the timber productivity of traditional forestry, and no research had been available on the ecology of old growth forests in the Pacific Northwest, many of the legal decisions leading up the FEMAT process probably would not have been made.

 

B. Internal Science Values.

1. Definitions of important scientific problems often reflect underlying, but unproved scientific premises -- i.e., assumptions about the nature of the world. Thus silviculturalists are prone to state that they can produce any forest conditions they are asked to produce, whereas old growth ecologists are prone to assert that "forests are not only more complicated than we think, but more complicated than we can think." These statements reflect fundamental premises that are empirical as well as normative, but generally cannot be proved or disproved empirically (cf., Majone 1984).

2. Scientific disciplines and their practitioners embody purposes and values. One of the individuals interviewed in our research stated the point very trenchantly: "Obviously one doesn’t do silvicultural research just in order to figure out how trees grow. You do it in order to produce boards. ... [W]ildlife biologists didn’t learn all about population dynamics to just merely describe how populations disappear. ... Yeah, sure, I’d say all the sciences or professions around natural resources are built within a context and that context is to be able to provide X, Y or Z on a sustained basis over a very long period of time. I don’t think any of us have ever gotten into the business of doing natural resources either management or research to merely understand how the system jigs around. That’s not the satisfying part of it. The satisfying part of it is to produce information that can be used in conservation."

3. Dealing with the pervasive uncertainty in natural resource science frequently poses serious science and policy problems (e.g., Thompson 1986). Scientists often disagree strongly about the implications of uncertainty. In the National Acid Precipitation Assessment Project, for example, the early lack of a known physiological mechanism through which acid rain would damage trees was taken by some scientists as a reason to doubt the validity of the apparent correlation between acid rain levels and forest health; others drew the opposite conclusion. Their respective views were probably related to trans-scientific values and assumptions about nature. Whichever view one took, however, had major policy implications. And there was no "purely" scientific or fact-based reason for choosing one position over another. Similar patterns have long been evident in EPA and OSHA regulatory decision-making, where a continuing question has been whether discernible chemical or physiological changes that have not been linked to demonstrable clinical changes should be treated as health effects (e.g, Majone 1984; Salter 1988; Jasanoff 1990).

4. Some of the most important work in natural resource science involves developing unifying concepts, or "master metaphors." Examples include "ecosystem integrity," "sustainability," "community stability," "social resilience," "ecosystem health," and the like (Meidinger 1995; cf., Botkin 1990). Because they help organize data which are otherwise effectively infinite and unintelligible, such concepts may be essential to any kind of analysis. However, they also contain normative implications about appropriate policy: e.g., do that which preserves ecosystem integrity or community resilience. Many modern natural resource science efforts involve intensive discussions and negotiations about unifying concepts, almost certainly in part because of their policy implications.

5. Aside from master metaphors, what data are gathered and how they are analyzed will sometimes significantly affect how the available policy options are defined. In the Pacfish case, for example, the statistical distribution of data led to statistical/distributive standards and guidelines for habitat, rather than more site-specific ones. This in turn may lead toward different ways of thinking about both policies and management strategies over time.

6. Many modern natural resource science problems operate on such large scales, involve so many variables, and require integration of so many disciplines that they may by nature be inconsistent with deterministic causal models often assumed to typify science. Instead they may systems models in which the scientists are participants as well as observers, and in which science is part of the system being studied. Thus, making decisions for science may in the long term actually amount to making decisions for the system being studied. Henry Regier (1993) has captured this situation most effectively with the diagram presented in Figure 3.

 

 

C. The External Effects of Science.

1. New knowledge and new conceptualizations can affect people’s attitudes toward aspects of their world, and are often produced in part for that reason. Thus, "new knowledge of harmful substances in the food we eat, the water we drink, and the air we breathe exacerbates our sense of the hazards of living." (Short 1990) Similarly, there is reason to believe that concepts such as ecosystem integrity and sustainability can serve to alter how people understand the implications of alternative choices, and are developed with those issues in mind. Moreover, although scientists cannot control how their work is used, there is no reason top think that science is "better" when it pays no heed to how its products might be used than when it does (Weiss 1984).

2. The difficulty of the advocacy question extends beyond the confines of scientific practice. Some of the individuals interviewed in our research stated that scientists should not advocate particular policies even in their capacities as private citizens, whereas others expressed the view that they had a right and possibly a duty to advocate specific policies as citizens. However, it is also arguable that scientific knowledge can impose moral and political duties upon those who possess it. The best-known example is that of nuclear physicists during and after World War II. After their research was used by policy makers to drop atomic bombs on Hiroshima and Nagasaki, widespread agreement emerged that their knowledge imposed a duty upon scientists to try to prevent further military use of atomic weapons. (e.g., Reville 1975).

 

IV. CONCLUSION

For the reasons described above, science intrinsically involves advocacy and values. To proceed as if it does not requires either ignoring the realities of science or hiding them. Neither response is desirable. At a minimum, when it is operating under impossible mythological standards, science will always be vulnerable to exposes. Perhaps more importantly, the real policy potential of science is unlikely to be realized if the discussion relies on a false picture of its practice and potential.

Accordingly, a more promising strategy is to pursue a serious discussion among scientists, delegated decision-makers, and citizens on what kinds of advocacy are acceptable or desirable among scientists, and under what conditions. With luck, that discussion may currently be underway. Although we would not venture to predict its outcome at this stage, we will venture several sideboards.

First, there will not be single set of rules delineating the appropriate role of scientists in all cases. Our research finds great variability in the science-intensive management and policy situations, in the nature of science involved, and in the role of scientists. These diverse situations will not yield a single set of guidelines, although they might yield a set of workable and acceptable patterns

Second, ongoing interactions between scientists and managers can lead to the accountability expected by placing a research branch within an administrative agency. However, realizing this potential requires protecting the critical capacity of scientists to question and debate current management practices.

Third, managers must become able and amenable to "thinking more like scientists" because few situations ever approximate the stable social, political and ecological situation assumed for management prescriptions to be unquestioningly applied as routines. Indeed, current efforts toward adaptive management clearly embrace the essential role of both management specialists and research scientists. This is a promising development which must be supported by administrators, especially science administrators since they will most likely worry about the slippage of research into management problem-solving.

Fourth, it is essential that science be understood as more than producing a "bucket of facts." Science transforms how society thinks about, values and acts toward natural resources. As such, scientists shape the nature of the world they study and thereby affect its future. The scientific study of biodiversity has transformed the global debate regarding the value of forests, and changed how people act as a result, and thus changed the nature of the forest.

Fifth, action very often precedes thorough scientific analysis. For this reason, scientists, managers, and policymakers should both treat management changes as experiments (Campbell 1970; Lee 1993) and constantly interrogate one another regarding possible pitfalls or unanticipated outcomes of possible courses of action. It is this new relationship that is so promisingly, if uncomfortably, emerging from science-policy assessments.

VI. LITERATURE CITED

Antypas, Alex and Errol Meidinger. 1995. "Science-Intensive Policy Disputes: A Review of the Literature." Report prepared for the People and Natural Resources Program, USFS-PNW Research Station, Seattle, WA.

Bacon, Francis. 1620. Novum Organum. 1620. (numerous English editions).

Botkin, Daniel B. 1990. Discordant Harmonies: A New Ecology for the Twenty-First Century. New York and Oxford: Oxford University Press.

Campbell, Donald T. 1969. "Reforms as Experiments." American Psychologist. 24:409-29.

Descartes, Rene. 1637. Discourse on Method. (available in many English editions)

Fuchs, Steven. 1993. "Positivism is the Organizational Myth of Science." Perspectives on Science. 1:1-23

Hays, Samuel. 1959. Conservation and the Gospel of Efficiency. Cambridge, MA: Harvard University Press.

Jasanoff, Sheila. 1990. The Fifth Branch: Science Advisors as Policymakers. Cambridge, Mass. and London, England: Harvard University Press.

Kaufman, Herbert. 1960. The Forest Ranger: A Study in Administrative Behavior. Baltimore, MD: Johns Hopkins University Press.

Latour, Bruno. 1987. Science in Action: How to Follow Scientists and Engineers Through Society. Cambridge, MA: Harvard University Press.

Lee, Kai. N. 1993. Compass and Gyroscope. Washington, DC and Covelo, CA: Island Press.

Majone, Giandomenico. 1984. "Science and Trans-Science in Standard Setting." Science, Technology and Human Values. 9:15-22.

Meidinger, Errol E. 1995. "Perspectives on the Role of Science in Natural Resrources Policy." Paper presented to the E1 and E5 Technical Sessions. Society of American Foresters, November 1, Portland, Maine.

Merton, Robert K. 1973. The Sociology of Science. Chicago: University of Chicago Press.

Peffer, E. Louise. 1951. The Closing of the Public Domain: Disposal and Reservation Policies, 1900-1950. Stanford: Stanford University Press.

Popper, Karl R. 1961. The Logic of Scientific Discovery. New York: Basic Books.

Regier, Henry. 1993. "Managing on the Way to Ecosystem Integrity." Windsor Ontario: International Joint Commission Tripartite Meeting. October 22.

Revelle, Roger. 1975. "The Scientist and the Politician." Science. 187:1100-1105.

Robinson, Glen O. The Forest Service: A Study in Public Land Management. Baltimore and London: Johns Hopkins University Press.

Salter, Leora. 1988. Mandated Science: Science and Scientists in the Making of Standards. Kluwer Academic Publishers: Dordrecht, Boston, London.

Shiff, Ashley L. 1962. Fire and Water: Scientific Heresy in the Forest Service. Cambridge, MA: Harvard University Press.

Short, James F. 1990. "Hazards, Risks, and Enterprise: Approaches to Science, Law and Social Policy." Law and Society Review 24:179-198.

Thompson, Paul B. 1986. "Uncertainty Arguments in Environmental Issues," Environmental Ethics. 8:59-75.

Wiebe, Robert H. 1967. The Search for Order 1877 to 1920. New York: Hill and Wang.

Weiss, Janet A. and Judith E. Gruber. 1984. "Using Knowledge for Control in Fragmented Policy Arenas." Journal of Policy Analysis and Management 3(2):225-247.