Hybrid regimes of knowledge? Challenges for constructing scientific evidence in the context of the GMO-debate

Over the last two decades, there has been a remarkable shift of attention to the scientific and political fundamentals of the precautionary principle. The application of this principle has become a main strategy of coping with the different forms and problems related to non-knowledge. Thus, societies are increasingly confronted with the challenging and hitherto unresolved problem of political and technological decision-making under conditions of diverging framings of non-knowledge. At present, there seems to be no generally accepted scientific or institutional approach. This is why the fundamental question of how different scientific actors define and construct evidence is not answered yet. Hence, this paper is based on the consideration that the conflicts in risk policy concerning genetically modified organisms (GMO) depend on the unresolved conflicts about the diverging scientific strategies and structures of evidence-making between the epistemic cultures involved. Thus, this study investigates two questions: (1) do the epistemic strategies of evidence-making differ systematically with the scientific actors involved in the GMO-debate? (2) What consequences emerge considering institutionalized procedures of decision-making?

This article is based on a secondary analysis of findings and perspectives reported in the literature and on the methods of qualitative social empirical research, i.e., interviews with experts. A total number of 34 interviews were conducted to explore the different strategies of handling non-knowledge and constructing evidence. Actors from science, administration, business and NGOs were interviewed. In this way, typical epistemic cultures can be described. An epistemic culture is the constellation of methodological strategies, theoretical assumptions and practical-experimental settings which define in every speciality the ways how we know what we know.

There are two main results. Firstly, it was worked out that the epistemic cultures involved in the GMO-debate use rather distinct strategies to define non-knowledge and to classify evidence. There are three types of constructing evidence, which correspond to different types of epistemic cultures. Secondly, the findings imply that the intensity of the conflicts in risk policy fields like the GMO-debate is due to a lack of knowledge politics. Usually, knowledge politics is restricted to the design of institutional procedures to compile knowledge provided by experts. The institutional setting of risk analysis and risk management is based on the premise of strict separation between knowledge and power. However, inadmissible mixing-up of knowledge and power is observable.

It seems that non-knowledge leads to an epistemic no man's land, and, hence, hybrid regimes of knowledge emerge. These regimes are hybrid with respect to the unclear and not explicitly reflected strategies of evidence-making. By lacking of knowledge politics, this situation opens up 'windows of opportunity' for actors with special interests in risk policy fields like the GMO-debate. Therefore, there is a difference between the visible institutionalized structures of risk policies and the rather invisible hybrid regimes of knowledge. Structure and scope of expertise have to be reflected and new instruments of knowledge politics have to be designed.

Different epistemic cultures can be qualified by describing their particular strategies of evidence-making. To solve the conflicts between these strategies, a meta-expertise is needed. Besides the institutionalized settings of knowledge politics, the underlying hybrid regimes of knowledge have to be identified.

The concept of epistemic cultures and their strategies of evidence-making should be investigated more explicitly with respect to other risk policy fields The analysis of hybrid regimes of knowledge should be deepened by looking at the complex interactions between institutional, discursive and practical rules affecting risk assessment.