The procedure used by Dutch, European and global authorities when granting licenses for genetically modified organisms is a flawed one. Although applicants are required to provide information about the dangers of the newly developed biotechnology to man and the environment, they themselves decide to some extent what information is relevant. This has been demonstrated by an analysis carried out by philosophers at Leiden University as part of a project funded by the NWO¹s Council for the Humanities. The applicant who is naturally an interested party is therefore in a position to influence the outcome of the approval procedure by deciding that certain information is irrelevant to the risk assessment.
The explanatory notes accompanying the questions on the current (Dutch) application form show that the government¹s decision whether or not to grant a license is based on the information provided by the applicant. It is stated, for example, that "synthesis of all the information provided in this application form is intended to result in a complete analysis of the risks." The questions which a biotechnology company is required to answer are phrased in general terms. For instance, applicants are asked about differences between the modified plant species and the original species and whether they expect hereditary changes to be spread within the environment. By formulating the questions in this way, the body awarding the licenses places the responsibility for the provision of information in the hands of the applicant.
The Leiden philosophers say that licenses can only be awarded in a responsible manner if the authorities have a list of questions which are relevant in assessing the dangers to man and the environment posed by modified organisms. As a start, they have drawn up a definition of the hazards posed by a modified organism, according to which a danger exists if "it carries an agent P which can produce an effect Q which is considered undesirable in context R on an affected item S by means of a mechanism T in an environment X as a result of application Z".
By making use of these categories, the relevant questions and the arguments for their relevance can be listed in a systematic manner. In the case of an agent, for example, the following question is in order: To what extent can the transgene survive without the genetically modified organism? The argument for the relevance of this question is that the transgene may constitute a danger by being transferred to other organisms.
The relevance of questions needs to be discussed by such parties as ecologists, molecular biologists, license applicants and the bodies granting the licenses. This would minimise unexpected environmental risks. In the past, the environmental dangers associated with DDT, CFCs and some synthetic chemicals have mainly been discovered by alert individuals and not by research institutes and companies. According to the philosophers, expert discussion in such leading periodicals as Science and Nature has to a large extent involved artificial controversies because the matter of the relevance of certain research questions was hardly considered. The omission in the bureaucratic licensing procedure is also to be found in the European Union¹s Directive 90/220/EEC and In the Familiarity policy pursued by OECD. The philosophers have now passed on their findings to the Commission on Genetic Modification (COGEM), the body which advises on license granting in the Netherlands.
The book ŒHazard Identification of Agricultural Biotechnology: Finding Relevant Questions¹ (reviewed in Science Vol. 286, 418-419, 15 October 1999) is available from T +31 30 273 1840, E-mail i-books@antenna.nl. Further information: Dr. Ad van Dommelen (Houtsma en Van der Schot) T +31 20 623 9480, F +31 20 638 8771 E-mail hout.schot@inter.nl.net