Better safe than sorry: : Applying philosophical methods to the debate on risk and the precautionary principle

Abstract: The purpose of the present thesis is to apply philosophical methods to the ongoing debate of the precautionary principle, in order to illuminate this debate. The thesis consists of an Introduction and five papers. Paper I con-cerns an objection to the method of conceptual analysis, the Charge from Psychology. After a brief characterisation of conceptual analysis, I argue that the Charge from Psychology is misdirected. In Paper II, the method of conceptual analysis is applied to the concept of precaution which is ana-lysed in terms of precautionary actions. The purpose is explicatory. A definition involving three necessary and jointly sufficient conditions is proposed, and the implications of this analysis for the debate on the pre-cautionary principle are discussed. Paper III attempts to provide an ana-lytical apparatus which may be used for finding improved formulations of the precautionary principle. The approach is lexicographical. Several exist-ing and possible formulations of the precautionary principle are examined, and four common elements and a common structure of the precautionary principle are identified. It is suggested that the analytical apparatus pre-sented can be used in negotiations of the precautionary principle. Paper IV questions the soundness of some arguments against the precautionary prin-ciple. Five common arguments are discussed and rejected. In Paper V, two of these arguments are further discussed. I argue that an attempt at rejec-tion of the precautionary principle delivered by John Harris and Søren Holm is unwarranted, because their arguments against it are based on in-terpretations of the precautionary principle that ignore context. Paper VI deals with the idea of de minimis risk. After a discussion of the distinction between disregarding a risk and accepting it, I examine one way of deter-mining how small a risk ought to be in order to be disregarded, namely the use of natural risk levels as benchmarks. I argue that this approach fails, even if the distinction between what is natural and what is not natural can be upheld.