Schwerpunktthema Diskursive TA-Verfahren

by René von Schomberg

The issue of the deliberate release of genetically modified organisms (GMOs) into the environment is perceived as a policy-problem with manifold aspects which could only be dealt with by appealing to science, manageability and social conventions. This threefold appeal to fundamental institutions of society implied a threefold reduction of the problem. The policy process for the release of GMOs into the environment is the focus of this paper and it is argued that a discursive policy process is needed to achieve an integrative, non-reductionist approach to the problem. I will argue that only a discursive policy process can overcome the problems of decision making in the context of uncertainties and I will here draw upon the experience gained during a EU funded study.(1) The main conclusion is that we have to acknowledge the logic of an uncertainty-based policy process, distinct from traditional risk-based policies, which follows a different rationale and implies different social-political assumptions. From this point of view I will comment upon van den Daele's interpretation of the results of the technology assessment procedure at the Science Centre Berlin.

2. The Appeal to Science

The authoritative appeal to science underlies the assumption that we have confidence in the functioning of the scientific system, for example, that it can provide policy makers with reliable knowledge and adequate predictions which are needed for a manageable practice for which policy-regulations must set the framework. In the case of the deliberate release of GMOs, the usual confidence in science is problematic for two reasons. First, we have to deal with a trans-scientific problem, that is a problem that can be stated in the language of science but cannot be solved within the language of science. Our current knowledge does not provide us with the means to predict the ecological long-term effects of releasing organisms into the environment. So it is beyond the competence of the scientific system to answer such a question, despite the fact that competence is normally the basis for an authoritative appeal. In fact, science would not pose such a question to itself since there is no method to make this question researchable. Reasoned statements on this subject matter cannot go beyond theoretical speculation. The reason for an appeal to science is solely policy motivated: we would like to have the answer to this question for achieving a manageable practice. So we have a good reason to be suspicious if scientists are nevertheless prepared to provide us with some kind of answer to this question. We can reconstruct two kinds of 'answers' science has given us so far. The first answer came from one branch of science, where most scientists were biotechnologists, molecular biologists or microbiologists. They answered the question by acknowledging the trans-scientific problem and stating that the development of a testprotocol for identifying the risks of individual organisms would be an unachievable task (Brill 1985).(2) However, at the same time they argued that this is irrelevant knowledge since we can rely on the experience with traditional plant breeding practices, which differs, on their account, insignificantly from the practice of genetic engineering - only in so far that we now exactly know what kind of new genes we are introducing. Ecologists on the other hand down-played the trans-scientific issue, by saying that they could develop precisely the type of knowledge policy makers asked for by doing research on so called 'microcosms' (cf. Krimsky 1996)(3) needed to make predictions possible in terms of quantitative risk assessment (Tiedje et al. 1989).(4)

From a policy perspective both answers are unsatisfactory because a biotechnologist cannot address the problems in terms of safety or in terms of risk. They just rhetorically state that it would be an 'acceptable risk' (by appealing to the fact that we already accepted the risks associated with conventional plant breeding). However this does not give us an informed opinion on how to regulate individual cases, nor did it address the issue of a precautionary approach. Ecologists, on the other hand, underestimate the difficulties of the trans-scientific issue: the promise of providing a quantitative risk assessment in the course of microcosm-based experiments, and without conducting field experiments, cannot be fulfilled in the foreseeable future. Only if one fully appreciates the trans-scientific issue, one sees the dilemma for policy: allowing major field experiments might involve unknown environmental impacts. To impose too many constraints on these experiments, however, would imply that we will never gain knowledge on the behaviour of GMOs. In the next section we will see that this dilemma bounces back on the regulatory system we have in place: what did we learn from the field experiments conducted during the last decade?

3. Appealing to Manageability

Hans Bergmans, Secretary of the Commission on Genetic Modification (COGEM) in the Netherlands, does not agree with the familiar argument that the field experiments with GMOs have demonstrated their safe environmental use.(5) According to Bergmans, it has only shown that experiments have been planned carefully. The experiments did not have any environmental impact other than those expected (to our knowledge). Consequently, the field experiments did not teach us anything about the behaviour of GMOs. This conclusion changes the initially intended perspective on the 'step-by-step' procedure. Rüdelsheim, from the company Plant Genetic Systems (PGS), based in Belgium, also affirmed this change in perspective at a workshop held in May 1995:(6)

One could say so far, the 'step-by-step' procedure focused more on the safety of the 'step' to be taken, than on the preparation of future 'steps'.

In conclusion, I would argue that if we still think that it is necessary to gain knowledge on the behaviour of GMOs we have to do something other than reviewing applications within the current 'step-by-step' procedure since it cannot demonstrate the safe environmental use of GMOs.

According to Bergmans, we should now plan experiments with an intended environmental effect, in order to gain the necessary knowledge. Bergmans advocates that we should allow applications with GMOs with similarly manageable effects such as the accepted agronomic effects of conventional agriculture. These types of releases could yield information on the behaviour of GMOs. Stressing the fact that only an increased knowledge of basic natural processes can help risk analysis, he also claims that it would be useful to use GMOs in order to increase our knowledge of micro-organisms in the environment and suggests that genetic modification can be used for the 'tagging' of micro-organisms so that they can be followed in the environment.

The task for policy is to translate the precautionary assumptions of the legislation which is based on a 'case-by-case' and 'step-by-step' procedure, into a manageable practice that acknowledges these assumptions and makes a science-informed learning process possible. We have observed that Bergmans addressed the trans-scientific issue explicitly, but we must realize that he has not given us an answer to the sequential questions: What intended effects can be 'manageable', on the one hand, and provide us, on the other hand, with usable information on the behaviour of GMOs that would provide a basis for risk assessment? What intended effects will be acceptable effects?

These questions cannot be answered yet, since not only the appeal to science implies a reduction of the problem, the manageability criterion, imposed by regulatory policy on the practice of field experiments has produced another possibly reductionist position: manageability has been equated with planning safe experiments from which we cannot learn enough.

4. Reducing Acceptability to Social Conventions

Existing legislation does not provide regulators with normative standards to evaluate applications concerning the acceptability of their environmental impact. Without a normative standard, however, it is impossible to draw a valid conclusion on the acceptability of a product or a release. Therefore, regulators have to make normative assumptions which could render a certain conclusion acceptable. So far, the implicit strategy has been to make an appeal to a conventional norm, that is to say a standard which would be acceptable because one can be certain it is widely accepted and uncontroversial. What kind of standard would that be? The Dutch advisory committee made the following statements in the evaluation of the application of the company Plant Genetic Systems in June 1994:

Outcrossing transgenic characteristics will not cause a persistently negative impact on the environment [and] outcrossing the gene and its property male sterility ... will not lead to a persistently unacceptable impact of these relatives on the composition of varieties in natural vegetation.

To draw a conclusion on the acceptability of an impact, one has to use phrases with normative implications like 'negative impact' or 'unacceptable impact'. In this case, the advisory committee assumed that a conventional standard, and therefore a noncontroversial reference point, would be the 'natural situation' itself. It is assumed that so long as any impact would be an impact which could be counterbalanced by nature, which would allow nature to return to its original situation, it would be an 'acceptable impact'. Generally, this conclusion, which at first glance seems quite uncontroversial, implies that any process or impact caused by releases or new agricultural practices would be acceptable if we found that such a process or impact would be an instance occurring in nature itself. Indeed, advisory committees came to the conclusion that herbicide-resistant genes, for instance, are widespread in the natural environment and that, therefore, a possible spread of these genes caused by manmade varieties would be an acceptable phenomenon, comparable with existing natural processes.

However unproblematic this appeal to a conventional norm seems to be, it soon runs into difficulties when one tries to apply this normative reference point in diverse cases over time. The assumption we make by its application is that we have a full understanding of natural processes. Now, in the case of the ecological impact of organisms introduced, we do not have such a body of knowledge. Our perception of nature changes over time and, for instance, up to some years ago we believed that a thing like 'gene flow' is not a natural phenomenon (and therefore unacceptable), but now we have found that it occurs in nature as well, which would turn it into an acceptable impact in case human practices would cause identical phenomena. So, our further analyses turn our 'convention' into a transformable normative reference point, which depends on (and evolves synchronically with) the historical change in our perception of nature.

Do we want environmental policy to be dependent on such standards? Regulators are now forced to study nature if they are to apply this standard consistently. Indeed, this is current practice to some extent. The assumption has always been that such a study would probably yield information that would eliminate the concept of hypothetical risk. Secondly, the standard would raise controversy if we were to say that anything happening in nature would be acceptable for human practices. Now we know that quite some natural events are unacceptable, otherwise it would not be possible any more to talk about natural catastrophes, precisely the kind of events some ecologists think that might happen with an intensified, biotechnologybased agriculture. Here we face the classical naturalistic fallacy: we cannot derive valid normative conclusions from factual statements. Thirdly, although we came to the conclusion that we are dealing with a transformable norm, since it is dependent on our perception of nature, it was not the intention of regulators to create such a standard (although the standard is rather well received by industrialists, who prefer to speak about 'flexible' standards).

In the statement of the advisory committee, that something is acceptable because it will not have a persistent negative impact, it is implied that there is a stable natural composition of the natural vegetation, enabling the vegetation to counterbalance any impact by returning to its original state. The keyword was 'persistent', implying a normative view on nature, which is perceived as a stable business counterbalancing any impact over a period of time. This normative view of nature has indeed been a quite influential conception in ecology for a long time but is now being replaced by the views of modern ecologists who introduced more 'realityadequate', more 'dynamic' models of nature. Who is right, is still undetermined; however, it seems problematic to take normative assumptions about nature's balance as a point of departure for evaluating the acceptability of an environmental impact.

Fourthly, the transformable standard introduced is of course very likely to become problematic in the light of other standards which, in their own right, are also introduced as standards referring to conventions. One could refer to conventional agricultural practices, that is to say, anything that does not yield an impact substantially different from the impact of existing agricultural practices. One could even refer to norms that 'should' become conventions in the very near future, like the standard of sustainable development, which is a normative guideline for the Danish authorities to evaluate the acceptability of an impact. Which standard do we choose?

Finally, we go back to the question of assessing the risks of GMOs. Regulators have been forced, in the absence of standards, to invent normative reference points to say something about the acceptability of an environmental impact. The statements on acceptability appealed to conventions. Indeed conventions refer to acceptable norms. However, they do not explicate the rightness of such norms or standards for which one has to argue. In doing so, they moved away from the practice of risk assessment to general statements on the acceptability of environmental impacts which can neither be defined in terms of risk nor in terms of safety standards. Since 'risk' presupposes a standard of acceptability, the regulatory system is not focused on identifying risks but rather on identifying uncertainties. We can distinguish between risk-based regulation (which applies to chemical substances) and uncertainty-based regulation for GMOs (see table 1.). The table shows, in accordance with our discussion above, that we have an uncertainty-based regulation in place whereas regulators and political actors often justify this type of regulation in terms of a risk-based regulation. However, no one at this point can either justify how to translate uncertainty to risk, or justify how to translate normative reference points to definitions of harm. The authorities in the UK claim to have a model for risk assessment in the framework of risk-based regulation. We have fully explained this model for the case of herbicide resistant oilseed rape. (cf. Levidow et al. 1996).

Table 1: Characteristics of regulatory systems

On the European level it became unavoidable to go beyond discussing safety issues without acknowledging that this is current policy. The unarticulated shift from risk-based regulation to uncertainty-based regulation needs a new justification since the vocabulary of a risk assessment model is inappropriate for current practice. Open discussion on transformable standards and the justification of an uncertainty-based regulatory system is hindered by the EC directive which restricts policy makers to the matter of scientific aspects of safety issues.

5. An appreciation of the Technology Assessment procedure at the 'Wissenschaftszentrum Berlin'

Our analysis of the assessments within the existing policy-procedures arrives at a similar conclusion as the TA procedure at the WZB in regard to one important aspect.(7) The plausibility of the occurrence of possible hazards of transgenic crops is less controversial than the use and selective choice of transformable standards, such as the normative reference to conventional agricultural practices, sustainable development, the natural vegetation or biodiversity. In so far I agree with one of van den Daele's major conclusions that these standards are in need of further (political) deliberation. However, following the line of our analysis I counterclaim his statement that transgenic crops 'involve no special risks' ('keine besondere Risiken'). Our conclusion is that these crops may involve special hazards ('besondere Gefahren'). Below I will try to substantialise this claim.

Both the TA procedure at the WZB and the deliberations within the existing policy-procedures have been focused on single cases. The focus on a single case has good reasons from a policy perspective. As I have summarized above the general scientific debate on the ecological effects of GMOs is inconclusive: in fact ecologists and biotechnologist base their prospective statements on assumptions and models which are all plausible to some extent but are unreconcilable at the same time (cf. von Schomberg, 1995). We only know that the contradicting models can not both be true in the long run. The persisting lack of consensus on the general level can be overcome by the practical policy decision to shift the discussion to specific individual releases. The counterfactual prospective claims on the possibility of the occurrence of hazards can not be easily demonstrated by ecologists in a specific case. Predictive ecology is in its infancy: A study of the specific releases should first yield the type of data on which ecologist could base their claims. In most of the cases this results in assessments that ultimately only appeal to the bias of the biotechnologist that the consequences of the releases can be foreseen on the basis of the familiarity principle (familiarity with the inserted gene and familiarity with the crop = safety). The familiarity principle is scientifically speaking incorrect but it is applied by biotechnologists because of its demonstrated practicality. This practicality simply overrides the need for new ecological data in individual cases. Although this is an unavoidable paradox within the existing regulation, one could reasonably argue that, indeed, the policy-procedure has resulted in a reduction of uncertainty for individual cases.

Van den Daele goes further and claims, that the TA procedure at the WZB, through its discursive nature, has demonstrated the incorrectness of the general model of the ecologists, once it is applied to an individual case:

Table 2: Policy issues for risk- based and uncertainty-based regulation

"Im TA Verfahren gab es (..) nicht einen endlosen epistemischen Diskurs zwischen Ökologen und Molekularbiologen, sondern eine sachliche Klärung mit einem eindeutigen Ergebnis" (8).

If I understand van den Daele correctly, the extra-quality and achievement of the TA procedure - in comparison with the general scientific debate which is also discursive in nature - lies in the "sachliche Klärung" of disagreements derived from the general scientific debate and applied to a specific case. Indeed, one should acknowledge such results as possible achievements of discursive TA-procedures. It is incorrect to suggest, however, that the TA-procedure allows anything more than provisional conclusions in individual cases. TA-procedures can yield a reasonable provisional consensus in individual cases, but do not remove the lack of scientific knowledge, nor overcome the epistemic differences among scientific disciplines. It is also incorrect to suggest that these provisional conclusions on individual cases allow us to say something substantial on the general level of the "risks" of gene-technology as such. The multi-facetness of the technology (virus-resistance, insect-resistance etc.) do not allow any general conclusion. The persisting disagreements, among scientists re-enters the debate as soon as the question emerges concerning the cumulative effects of 'thousands' of releases or the unassessable long-term effects of a variety of releases which may cause multiple-herbicide resistance. This debate lacks a functional equivalent for "Grenzwerte" from conventional risk-based policies. No one has suggested a solution for the problem how to define the borderline for the environmental pressure caused by numerous releases. We assume to have answered this question before we have reached an unacceptable state. The impossibility to foresee the long-term and the cumulative effects of releases, together with the lacking ability to define exactly how they would look like, forces us to continue with a precautionary case-for-case approach and invest in the monitoring of releases. In this context the TA-procedure at the WZB has been an important experiment which should be interpreted in terms of uncertainties and provisional conclusions on individual cases rather than in terms of risks of gene technology. In fact, TA procedures could result in the transformation of the uncertainty-issue by extending the discursive procedures to the level of the transformable standards which has proven to be more controversial than all plausible environmental effects from any release. My plea would be to fully embark on a course towards a discursive policy encompassing an integrative approach to the subject matter that would facilitate a flexible regulation, both effective and legitimate. However, to meet this objective we have to formulate procedural norms for this discursive process to make a fair and just outcome possible.(9) In table 2, I summarize the policy-issues related to risk-based regulation distinct from uncertainty-based regulation.

Notes

(1) This essay arises from the context of a study for the European Commission "GMO releases: Managing Uncertainties about biosafety", conducted by Les Levidow et al., Open University, UK. The results of this study have been published in the special June-issue of Science and Public Policy 1996.
For a recent update of the scientific debate, regulatory practices, case-studies and socio-political analysis, see: "Coping with Deliberate Release. The limits of Risk Assessment", edited by Ad van Dommelen, published in 1996 by the International Centre for Human and Public Affairs, Tilburg, The Netherlands.

(2) Brill, W.J. 1985: 'Safety concerns and genetic engineering in agriculture.' Science 227, pp. 381 - 384.

(3) See his contribution in van Dommelen, cited in note 1.

(4) For a thorough analysis of the debate between biotechnologists and ecologists, see my book: Der rationale Umgang mit Unsicherheit. Die Bewältigung von Dissens und Gefahren in Wissenschaft, Wissenschaftspolitik und Gesellschaft, Frankfurt am Main: Peter Lang, 1995. A summary appeared in: René von Schomberg (ed.), Science, Politics and Morality, Scientific Uncertainty and Decision Making. Dordrecht: Kluwer Academic Publishers, 1993.

(5) Qotations are from Bergmans' paper presented at the workshop: "Unanswered safety questions when employing GMOs", organised by the Coordination Commission Risk Assessment Research on 2-4 May, 1995.

(6) Qotations are from Rüdelsheim paper presented at the workshop: "Unanswered safety questions when employing GMOs", organised by the Coordination Commission Risk Assessment Research on 2 - 4 May, 1995.

(7) Daele, W. van den; Puhler, A.; Sukopp, H. et al.: Grüne Gentechnik im Widerstreit. Weinheim: VCH-Verlag 1996.

(8) "Objektives Wissen als politische Ressource". In: Daele, W. van den (Hrsg.), Kommunikation und Entscheidung. Politische Funktionen öffentlicher Meinungsbildung und diskursiver Verfahren.WZB-Jahrbuch, 1996.

(9) On this topic, see: 'The erosion of the valuespheres. The ways in which society copes with scientific and moral and ethical uncertainties.' In: René von Schomberg (ed.), Contested Technology. Ethics, Risk and Public Debate. Tilburg: International Centre for Human and Public Affairs, 1995.

René von Schomberg
Faculteit der Wijsbegeerte
Tilburg Universiteit
P.O. Box 90153, NL-5000 LE Tilburg
Tel.: + 31 13466-3018