Diagnosing and prioritizing uncertainties according to their relevance for policy: The case of transgene silencing

Abstract

Uncertainty often becomes problematic when science is used to support decision making in the policy process. Scientists can contribute to a more constructive approach to uncertainty by making their uncertainties transparent. In this article, an approach to systematic uncertainty diagnosis is demonstrated on the case study of transgene silencing and GMO risk assessment. Detailed interviews were conducted with five experts on transgene silencing to obtain quantitative and qualitative information on their perceptions of the uncertainty characterising our knowledge of the phenomena. The results indicate that there are competing interpretations of the cause–effect relationships leading to gene silencing (model structure uncertainty). In particular, the roles of post-transcriptional gene silencing, position effects, DNA–DNA interactions, direct-repeat DNA structures, recognition factors and dsRNA and aberrant zRNA are debated. The study highlights several sources of uncertainty beyond the statistical uncertainty commonly reported in risk assessment. The results also reveal a discrepancy between the way in which uncertainties would be prioritized on the basis of the uncertainty analysis conducted, and the way in which they would be prioritized on the basis of expert willingness to pay to eliminate uncertainty. The results also reveal a diversity of expert opinions on the uncertainty characterizing transgene silencing. Because the methodology used to diagnose uncertainties was successful in revealing a broad spectrum of uncertainties as well as a diversity of expert opinion, it is concluded that the methodology used could contribute to increasing transparency and fostering a critical discussion on uncertainty in the decision making process.