The contradictions articulated in the process of rethinking scientific and political rationality are addressed differently by colleagues working from various disciplinary perspectives and research traditions. What can be witnessed is a huge diversity of research problems, concepts, and approachesâeven conflicting episte-mologies. But what they have in common is their struggle with at least that part of the contradictions which is related to the rethinking of scientific and political rationality. This is what makes them valuable for exploring the nexus of knowledge, power, and participation.
Knowledge Utilization Studies
The topics addressed in political science build on a tradition among policy scientists and sociologists that can be referred to as knowledge utilization studies. The main questions for scholars working in this field are how to understand and explain knowledge use in public policy.
In many publications, the underlying assumption is that knowledge from the sciences (natural and social) is underused or not used at all. The Two Cultures theory (Snow, 1959; Caplan, 1979) explains this phenomenon by pointing out that science and policy represent different cultures that are hard to relate. From this perspective, knowledge and power do not easily merge. As to whether this is considered unfortunate, that depends on how one thinks of the social function of science. The enlightenment model boils down to a trickle-down theory of knowledge use, or âknowledge creepâ (Weiss 1981). In their universities and R&D labs scientists think and experiment; mature results are published in learned journals and books. Mediated by popularizing scientists or scientific journalists, they enter the world of administrators, public officials, and politicians. What they do with the scientific information is up to them. Hopefully it is enlightening; at worst it is ignored (perhaps at their peril). But neither situation is the concern of science.
Others have argued that knowledge is being used to the extent that the information provided fits in with the interests and values of the user. Rich (1991) asserts that policymakers have a limited interest in searching for information, as they are unable to handle new and often contradictory knowledge claims. The explanation for this phenomenon is provided by the concept of bounded rationality (Simon, 1957). Decision makers tend to ask for and use information from knowledge providers immediately linked to their own agency. Only if they sense that the knowledge they need is not available in their immediate network, or if they are uncertain, will they ask for knowledge from university and other research institutions outside their own agency. According to this view, the primary social and political function of knowledge is to confirm and strengthen already established positions and views.
This perspective, when taken to a macro level of analysis, fits in very well with a research-and-analysis-as-intellectual ammunition conception of knowledge use (Lindblom, 1968; Coleman, 1979), which has been labeled as the adversarial model (Wittrock, 1991). The core assumption of this model is that politics is a pluralist interest group struggle involving partisan mutual adjustment (Lindblom, 1965). Processes of partisan mutual adjustment work like a selection machine of scientific arguments for already fixed political stands. Each and every interest will mobilize its own science-based expertise to bolster its case. Policy analysts are like lawyers, and their business is advocacy. Policy analysis may enrich the quality of political debate, provided that all interests have their own expert. To the extent that controversies maximize the mobilization of expertise, they may be conducive to the utilization of knowledge. In the adversarial model, separate actors defend or strengthen their respective positions in the short run, while in the long run policy-oriented learning may result (SabĂ€tier and Jenkins-Smith, 1993). The concept of policy-oriented learning is based on the idea thatâbeyond political strifeâthe political community eventually learns how to best solve its problems.
Whereas Sabatier and Jenkins-Smith build upon the American pluralist model of policymaking based on advocacy coalitions between policymaking agencies and interest groups, other approaches to learning recast partisan mutual adjustment along social and culturally determined institutional lines, as a struggle for cultural (Douglas and Wildavsky, 1983; Rayner, 1991) or discursive hegemony. Political scientists and sociologists who work from a discourse perspective do not take an a priori position with respect to the kind of cleavages (based on interest, culture, or other) between competing policy and knowledge claims. They are more interested to empirically observe and understand the dynamics that shape knowledge for policy and policy for knowledge (Hajer, 1995; Wynne, 1994; Shove and Redclift, this volume). Therefore, these authors do not endorse general expectations or theories with respect to systemsâ (in)abilities for policy-oriented learning. Worth mentioning are those who, instead of focusing on policy learning, point attention to the phenomenon of social and political unlearning, especially Lindblom (1997) in his discussion of cognitive impairment. In the same direction, Ravetz (1993) discusses âsocially constructed ignoranceâ as an inevitable consequence of the structuring of debate and inquiry.
It should be noticed that adversarial, policy-learning, and discursive approaches might reflect a specific view on the merger of knowledge and power. This is in so far as they share the methodological observation that policies can be treated as hypotheses (Hoogerwerf, 1990), policy implementation as social experiments (Pressman and Wildavsky, 1971), and policy belief systems as Lakatosian research programs being progressive or degenerative in spawning solutions to policy problems (Majone, 1989). Furthermore, these research perspectives tend to broaden the concept of knowledge, which comes to include both scientific and practical knowledge.
There is broad agreement that scientific and practical knowledge may be improved by employing rationally defensible (and hence nonarbitrary) standards of assessment to evaluate the evidential merit of competing knowledge claims. Apart from this general commitment, there is little agreement on the character of standards of assessment that ought to be employed. Proponents of a view according to which the sciences and professions are seen as vehicles for simultaneously acquiring intellectual and practical knowledge (e.g., Campbell, 1988) tend to consider the serving of practical interests as something which may compromise and even corrupt the quality or validity of science. Here, validity is defined according to standards for making plausible causal inferences in the face of rival hypotheses. By contrast, proponents of a view according to which the sciences and professions are seen chiefly as instruments for acquiring practical knowledge (e.g., Schön, 1983) tend to view the process of serving practical interests as an opportunity for improving both the quality of the sciences and professions, and their capacity to solve practical problems. Here, quality is likewise defined in terms of standards for making plausible causal inferences; but reflective practitioners are generally seen to be more reliable sources of such inferences than persons nominally designated as âscientists.â Whereas many policy scientists appear to turn away from this problem of supremacy by stating that next to scientific knowledge, practical knowledge is also or equally important, some search for an episte-mologically and methodologically sound way of articulating the functions of both kinds of knowledge (Dunn, this volume).
So far, we have addressed some of the multitude of theories which explain the use and nonuse of scientific knowledge in terms of social, cultural, or political cleavages (for other models see Wittrock, 1991; Hoppe, 1999). There are also theories that tend to explain the relations of knowledge and power in terms of stages or cycles, for example the concept of the âlife-cycle of information,â or what Rich (1981a) calls the âknowledge cycle.â These cyclic frameworks, while they formally provide for feedback as well as feed-forward loops, are often quasi-linear by virtue of their common emphasis on chainlike activities that are irreversible. Environmental management frameworks tend to relate the need and use of knowledge to stages in policymaking, such as problem recognition, policy formulation, policy implementation, and monitoring.
Functional frameworks originate in the sociology and economics of knowledge applications (e.g., Holzner, Dunn, and Shahidullah, 1987; Holzner and Marx, 1979; Machlup, 1980). Interdependent knowledge functions, along with the social structures by which these functions are performed, are viewed as a complex social system of knowledge, or âknowledge systemâ for short...