FS H 94-319
Technology Assessment as a Political Experiment Discursive Procedure for the Technology Assessment
of the Cultivation of Crop Plants with Genetically Engineered Herbicide Resistance1
Wolfgang van den Daele
Berlin, August 1994
1 A slightly amended version of this text is contained in the documentation on the technology assess
ment procedure with its various indices. See van den Daele, W., "Technikfolgenabschätzung als politisches Experiment"; in van den Daele, W., Pühler, A., and Sukopp, H. (eds.), Verfahren zur Technikfolgenabschätzung des Anbaus von Kulturpflanzen mit gentechnisch erzeugter Herbizidresi
stenz, Heft 1, Wissenschaftszentrum Berlin, 1994, Discussion Paper FS II 94-301.
SUMMARY
From February 1991 to June 1993 some 50 representatives from industry, environmental groups, regulatory agencies and science cooperated in a technology assessment on genetically engineered herbicide resistant crops. The participants in the technology assessment proce
dure were involved in more than 10 days of controversial debate and analysis. They dis
cussed and evaluated nearly 20 expert reports which covered issues of risks and benefits related to the release of genetically engineered plants and the use of new herbicides in agri
culture.
The Science Center Berlin is publishing the reports and proceedings of the technology assessment. This paper is a shortened, English version of volume 1 of the collected materi
als. It presents the structure and history of the technology assessment. In particular, it describes how the participatory mechanisms were organised and through which steps conclu
sions with respect to the herbicide resistance technology have been derived from mostly con
troversial discussions among the participants. The description is substantiated by various documents in the appendices. The representatives from the environmental groups resigned from the technology assessment by leaving the final conference at which conclusions were to be discussed. This report and the proceedings of the technology assessment will enable the public to judge for themselves whether the procedure was fair and the conclusions warranted.
The German version of this report appears as Discussion Paper FS II 94-301.
ZUSAMMENFASSUNG
Die Abteilung "Normbildung und Umwelt" im Wissenschaftszentrum Berlin hat von 1991 bis 1993 ein Verfahren zur Technikfolgenahschätzung für den Anbau von Kulturpflanzen mit gentechnisch erzeugter Herbizidresistenz mitorganisiert und begleitet. In dem Verfahren wurden alle wichtigen Problemfelder der Herbizidresistenztechnik mit Hilfe von Gutachten untersucht und in kontinuierlichen Diskursen von den Beteiligten (unter Einschluß von Befürwortern und Gegnern der Technik) erörtert.
Die Abteilung ’'Normbildung und Umwelt" veröffentlicht die Materialien des Verfahrens.
Dieses Heft ist eine (gekürzte) englische Fassung des ersten Heftes dieser Materialien. In diesem Heft werden die Struktur und der Ablauf des Verfahrens zur Technikfolgen
abschätzung dargestellt. Insbesondere wird beschrieben, wie die Partizipation der Beteiligten organisiert war und wie man im Verfahren aus den oft kontroversen Diskussionen Schlußfolgerungen für die Beurteilung der Herbizidresistenztechnik abgeleitet hat. Diese Beschreibung wird durch eine Reihe von Dokumenten im Anhang ergänzt. Die Vertre- ter/innen der Umweltverbände haben das Verfahren zu Beginn der Abschlußkonferenz verlas
sen. Die Öffentlichkeit kann sich an Hand dieses Heftes und der zur Technikfolgen
abschätzung veröffentlichten Materialien selbst ein Urteil darüber bilden, ob das Verfahren fair war und die Schlußfolgerungen gerechtfertigt sind. Die deutsche Fassung ist als Discus
sion Paper FS II 94-301 erhältlich.
CONTENTS
1. Introduction 1
2. Transgenic, herbicide-resistant plants as the subject: technology-induced 2 versus problem-induced technology assessment (TA)
3. The central role of information as the basis for TA, expert competence 5
4. Selection of the subjects for the expert reports 7
5. Representative participation in the TA procedure 10
6. Fair allocation of resources 12
7. Steering of the procedure: participants, Coordinating Committee, 13 WZB working group
8. Loyalty vis-ä-vis the procedure 14
9. Issues considered: risks, advantages, alternatives to HR technology 16 10. Evaluation conferences: science-based discussions and process orientation 19 11. Discursive procedure, pressure for consensus and statements of dissent 22 12. Final conference of the procedure: Instructions for the final report 25 13. Some structural problems of participatory technology assessment 27
References 31
Appendices 33
I: List of reports commissioned in the technology assessment 33 II: Distribution of participants in the TA procedure and allocation of 36
resources (by institution and group)
III: Procedural steps in the technology assessment: conferences and 37 workshops
IV: Press releases concerning the resignation of the environmental 39 groups from the technology assessment (in German)
1. Introduction
Between 1991 and 1993 a technology assessment (TA) of crop plants with genetically engineered herbicide resistance (HR) in agriculture was undertaken in the Federal Republic of Germany. The procedure was organized by the Wissenschaftszentrum Berlin (Prof. Dr. Wolfgang van den Daele) in cooperation with Prof. Dr. Alfred Pühler, Institute for Genetics of the University of Bielefeld and Prof. Dr. Herbert Sukopp, Institute for Ecology of the Technical University Berlin. Financial support was provided by the German Ministry for Research and Technology. In the TA pro
cedure various angles to embark on new courses have been tried. The main key words here are: participation and discourse.
The approach taken in this TA is based on the assumption that TA should not merely be a forum of experts at which the state of knowledge on the possible consequences of a technology is presented and evaluated. TA should, in addition, be an "arena" in which the social conflicts related to the introduction of a new technology can be articulated and discussed in an exemplary manner.
Hence, participants in the procedure were put together so as to reflect all the interests and the positions of the on-going political conflicts over new technology and to include the declared advocates and critics of the specific technology under considera
tion. In this way, the disputes which normally take place outside TA (and only become really heated when the results of TA are made public) were built into the pro
cedure from the very outset.
The TA procedure was organised as a social process of ongoing communication amongst those present, in order to guarantee a dialogue between the representatives of controversial positions. In a series of conferences the participants were to define the scope of the study framework, to evaluate the results of the expert reports, and to dis
cuss any conclusions which could be drawn. The understanding was that such a communication would promote discursive forms of debate, increasing the likelihood that arguments from the two sides were heard, and critical appraisal of the respective positions was accepted.
Whether the procedure actually fulfilled these expectations is something which those involved will probably view differently. At the beginning of the final conference in June 1993, the environmental associations announced their withdrawal. It must be left to the observers of the procedure to judge this step. In the following, I have limited myself to a more detailed presentation of the procedure itself and the main
phases it entailed. The analysis of the TA procedure is the subject of further research of a study group at the WZB.2
2. Transgenic, herbicide-resistant plants as the subject: technology-induced versus problem-induced technology assessment (TA)
HR technology opens up new opportunities for chemical weed control. It makes the use of non-selective herbicides in farming possible. So far these herbicides could, in general, not be used in agriculture, because they attack not only the undesired weeds but also crop plants which are to be protected. This obstacle is removed when genes are transferred to crop plants which make them resistant to or tolerant of the non- selective herbicide.3
Herbicide resistance was chosen as the subject for the TA procedure because it seemed to be sufficiently relevant and controversial. Transgenic, HR plants are viewed as one of the projects in which the practical application of genetic engineering in agriculture was emerging. Their development was well advanced. Up to 1990 almost half of all release experiments involved transgenic, herbicide-resistant plants.4 A first commercial variety (imidazolin-resistant maize) was placed on the market by the US seed company, Pioneer, in 1992.
HR technology was quick to attract public criticism. The criticism was sparked off by the possible risks of the large-scale release of transgenic plants, and by the prospect of optimised and perhaps even more extensive use of chemicals in weed control. In many ways HR technology was and is still viewed as the paradigm of an undesired development conditioned by the prevailing economic and technical trends.5 That is the very reason why this technology has been made the subject of the TA procedure.
2 For details on the theoretical approach of this research and initial results, see van den Daele (1992), Bora/Döbert (1993), Bora (1993), Döbert (1994), van den Daele/Döbert (1994).
3 For the technical details and the various resistance mechanisms, cf. the expert reports of Böger (1994) and Broer/Piihler (1994), vols. 2 and 3 respectively of the documentation on the technology assessment procedure.
4 According to a study by Ciba Geigy Seeds up to the end of 1991, 169 field trials were carried out with herbicide-resistant plants or approved for that purpose (all outside Germany). Twenty-four different combinations (varieties/herbicide resistance) were examined. Of the 155 field trials in 1991, 56 (36%) involved transgenic, herbicide-resistant plants. In 1990 the figure was 47 out of 107 (44%); see Chasseray/Düsing (1992). In Germany an application for the first field trials with transgenic, herbicide-resistant plants was for 1994.
5 Cf. Meyer-Abich (1984, p. 249). Winter (1992, p. 389) picks herbicide resistance technology as a negative example, too. The German Government has decided not to provide public funding for any of the projects involving HR technology
It could be expected that the TA invited consideration of a broad spectrum of the problems of modern technology development, e.g.:
• the possible risks of transgenic plants,
• the toxicological and ecological effects of the use of non-selective herbicides,
• the future of genetic resources,
• the advantages and disadvantages for farming,
• economic and agricultural implications,
• the long-term safeguarding of world food supplies,
• the ethics of plant manipulation.
All the same, the scheme of the TA procedure remained somewhat conventional in that a so-called "technology-induced" TA was envisaged. The starting point is the emerging technological development, and the analysis focuses on the possible conse
quences of this development. The goal is to assess the technology and to determine the political actions which might be necessary to cope with that technology. Most of the studies, which bear the title of "TA" proceed according to this pattern.
The technology-induced approach was rejected by the critics of the HR technology at the beginning of the procedure. What was called for instead was a "problem-induced"
technology assessment.6 Such an assessment does not address the consequences of a specific technology but examines which social problem the technology is supposed to help to solve. It then compares the various ways of tackling the problem. In the case of herbicide resistance, the starting point would be the agricultural problem of weed control. A comparison would then be made of the solutions offered on the one hand by industrial, intensive farming and, on the other, by ecological farming. A compari
son of this kind, however, should not be restricted to strategies for weed control. It would have to take into account the overall production systems in which these strate
gies are embedded. The decisive questions would then be whether we really need HR technology and what kind of farming is socially desirable and ecologically acceptable.
"Problem-induced" TA allows the discussion of broad and fundamental political issues. However, it, in fact, inflates the problems of TA to problems of political planning in general: Which objectives should be pursued in society? How can they be
6 For the opinion of the Gene-Ethical Network on the TA procedure at the first conference in Loc- cum, Germany, February 1991, see Gill (1991).
achieved? What are the opportunities and what is the proper role for state control?
Whether scenarios of desirable futures should really be called "TA" may, in the final analysis, boil down to a battle of words. There can be no question that scenarios of this kind are a good thing. But it is not so clear that they are the only good thing.
Problem-oriented TA cannot substitute technology-induced TA. The latter is always implied in the former and a precondition for it. Had the participants come out in favour of extending the assessment of HR technology into a systematic comparison of industrialised agriculture and ecological farming as paths for agricultural develop
ment, none of the issues raised in the technology-induced TA procedure would have been invalid.7 Additional and corresponding questions would have had to be addressed for ecological farming, too. The programme of the TA would at least have been doubled.
The dispute about the proper TA approach is as old as the TA itself. The more TA is seen as a vehicle to revise in principle the dependence of modern societies on techno
logical dynamism the broader the horizon of problems to be considered. Thus TA has regularly proceeded from the analysis of the consequences to the analysis of the ori
gins of the technology in question, moving from control to design and from technical options to the social needs. Equally regularly, however, political and pragmatic con
straints imposed a return to classical TA oriented towards the control of the conse
quences of a technology. In reality, we will have to work towards the accumulation of and cooperation between all the various approaches.
For our TA procedure a switch to a problem-oriented approach was rejected for financial reasons at the first conference. It thus remained an open question whether such a switch would in principle have been acceptable for all participants. As a mat
ter of fact, the organisers of the TA had committed themselves to the conventional scheme of a technology-induced TA. This commitment was not up for discussion. The necessary resources to extend the assessment into a comparative system analysis of paths of agricultural development were not available.8 In order to guarantee that the alternative of ecological farming was represented in the procedure, a special expert report on weed control in ecological farming was invited. The environmental groups
7 The evidence, too, that a problem can he solved with traditional, technological resources does not mean it is no longer necessary to examine new technologies. The new technology could bring with it a major improvement. Studies of alternatives place symmetric demands on new and old technolo
gies—unless there are reasons why new technologies would in principle be less favourable than old ones.
8 It is not clear whether a renewed application to the German Ministry for Research and Technology would have been successful. An analysis of "agricultural development paths" had been recently commissioned by the German Bundestag. Cf. Enquete Commission (1990).
have (temporarily) accepted this arrangement. The working programme of the TA procedure has been approved in consensus and ceased to be a subject of dispute until directly prior to the final conference.
3. The central role of information as the basis for TA, expert competence
Political evaluation is the final stage in every TA. The essential strategy of the TA is, however, investigation oriented towards factual information. Evaluations are based on reasons which have empirical references, that is, they refer to statements on causal links, facts, phenomena which (at least in principle) are verifiable of refutable.
Whether unintended metabolic changes justify a ban on transgenic plants or whether increases in yield through the use of herbicides are useful, is a value judgement.
Whether metabolic changes do occur and whether increases in yield can be expected is something we can know. And even if we determine that we do not or cannot know, we still refer to the domain of knowledge. Establishing the limits to knowledge is a matter of knowledge.
Information orientation of TA is the political basis for the relative consensus, which supports to the institutionalisation of TA in our society. Were political interests the only aspect that really mattered, then TA would be a waste of both time and money.
The interests are, in general, known in advance. The important point of a TA is that interest which participants pursue when they enter the TA procedure must be sifted through the filter of information orientation and, so to speak, expose themselves to the risk of information. The very idea behind the TA is that it involves the possibility that handy claims, by means of which one can beat the drum for a technology in pub
lic or provoke resistance to the same technology, can be shown to be unprovable, poorly justified or simply wrong. That does not mean that political interests or goals can themselves be "refuted". But they can lose their favourite legitimation. To justify them, new reasons have then to be provided. This risk can only be circumvented by circumventing TA altogether. Once you participate, you are automatically exposed.
Information orientation imposes limits on the function of the TA procedure to provide a forum for the political conflict over technology. Despite this or perhaps because of this (depending on the political stance adopted), it has been accepted as the working basis by all those involved in the HR procedure. The Gene-Ethical Network (a lead
ing group of activists campaigning against genetic engineering) declined to participate exactly because the TA procedure seemed incompatible with the straightforward goal of political mobilisation of the general public. This abstention was declared to be a
matter of political "division of labour" among the critics of the technology. It was not (at that time) considered a fundamental rejection of the TA. It is very possible that the virtualisation of political aspects through the information orientation of the TA, was only reluctantly accepted and viewed as a concession by the social move
ments. For industry and many scientists it was an essential precondition for their participation. They would have declined to just talk politics. Compromise was necessary. Within the framework of voluntary TA, the conditions for participation must be symmetrical. Criteria can legitimately claim that structural disadvantages the oppositional groups in society have to face must not affect their position in the proce
dure. Fair and equal treatment must be guaranteed. These groups cannot, however, expect that disadvantages they suffer in other contexts be compensated for by prefer
ential treatment in the TA procedure.
There was consensus in the HR procedure that expert reports were to be a main ele
ment in the TA. However, there was also general agreement that the expert reports should not be the only result of the procedure. They were to provide the material for the political evaluation of the HR technology. Furthermore, it was agreed that issues of information were first and foremost the competence and responsibility of the experts from the pertinent scientific and technical disciplines. The, in some cases, very fundamental political criticisms that has been waged against expert knowledge and expert cultures both in the social movements and in the social sciences during the last 20 years did not play any role in the TA procedure.
The responsibilities of the experts were probably less problematic in the TA procedure because greater transparency was guaranteed here than in other contexts of scientific advice for politics. The role of the experts was modestly defined and exaggerated claims or "technocratic excesses" were generally excluded. The inherent limits to expert knowledge were a constant subject of discussion in the procedure. Every expert report had the explicit task of presenting what was known about a problem, what could be known and what was not known. The bringing together of experts with different political convictions guaranteed that the unavoidable "softness" of expert opinions and scientific controversies, where they existed, could be brought to light.
The participants in the procedure agreed that forecasts on the basis of theories are never absolutely certain, that the examination of empirical claims depends on the method adopted and that science can only depict complex realities in a limited man
ner.9
9 For example, nobody claimed that scientific tests could really prove that a technology was safe.
Everybody admitted that one can only examine whether specific indicators for a risk are given or
It was equally clear in the procedure that experts are not truly "disinterested", they do have their interests and make value judgements. The experts were even explicitly asked (although in most cases in vain) to formulate clear evaluations, which were to trigger political discussion. To that extent, the supposedly political "neutrality" of the experts was not an issue. This does not at all mean that the reverse conclusion was drawn that experts were merely the representatives of interests. The participants in the procedure, as a group, were able to distance themselves from the value judge
ments of the experts, to differentiate between facts and evaluations, and to establish what the experts knew and what they wanted in political terms or thought was right.
Thus, the politicalisation of the experts in the TA procedure was within bounds.
Legitimate room for political judgements was provided in the selection of the ques
tions to be addressed in the expert reports and in the evaluation of the findings. The findings, themselves, were supposed to be knowledge not politics. They were a mat
ter of insight, not of interest.
4. Selection of the subjects for the expert reports
The selection of the subjects of and the individuals for the expert reports which were to be prepared in the TA procedure was dealt with on the first conference in 1991.
To this end, a WZB working group prepared a provisional programme structure. The programme proposal was the result of numerous preliminary discussions with poten
tial experts conducted over a period of eight months. The declared goal was to cover, if possible, all the problems associated with HR technology in scientific and public communications. The potential experts were invited to present their proposals for reports for discussion at the first conference.
At the conference the programme proposal was criticised from various sides. On the one hand, the broad range of topics was criticised, which often did not reveal a spe
cific link to HR technology. This argument was in particular raised against tine planned studies on the problems of non-selective herbicides. On the other hand, the spectrum of subjects was criticised as being too narrow. Additional reports were called for by the critics of HR technology, e.g. on soil erosion, on the lack of enforcement of application restrictions for herbicides, on the patenting of genetic
not. That the technology is seen as being sufficiently safe when this examination turns out negative is a political judgement.
resources from the Third World by companies from industrialised countries, on the status of the interests of industry and the environmental associations.
The organisers defended the proposed programme structure. Final agreement was not reached during the conference itself. In particular, the additional reports requested by the opponents to HR technology remained controversial. The call for a second pre
paratory conference at which all participants of the TA could continue to negotiate further details of programme and procedure was rejected because of the limited resources available and the time schedule envisaged for the TA. Further discussion of the study framework (scoping) was delegated to the Coordinating Committee (see below). In two meetings the committee drew up a final programme proposal which attempted to reconcile as far as possible the different demands. The proposal was presented in written form to all participants for comments and was then taken as accepted as no further objections were raised. (The details of the research programme may be gathered from the list of reports issued from this technology assessment; see appendix I.)
The Coordinating Committee, together with the experts participating in the TA, developed three scenarios for the application of HR technology. The scenarios set out preconditions to be taken into account in the expert reports on the ecological and agri
cultural impacts of non-selective herbicides, thus enhancing comparability of the findings. They determined which herbicides, cultures and areas of HR application should be considered in the reports. They distinguished three grades of implementa
tion:
Scenario 1 (medium implementation) Application of HR technology in only one mem
ber of a crop rotation. Sugar beets, potatoes, and rape are to be taken into account, each in a typical 3-member crop rotation.
Scenario 2 (high implementation)
Application of HR technology as in scenario 1 and, in addition, in maize in permanent cultiva
tion.
Scenario 3 (very high implementation) Application of HR technology in all members of a crop rotation.
The final programme structure contains a series of decisions which were of strategic importance for the TA procedure.
A strict limitation to the specific problem aspects connected to HR technology was not accepted. This was appropriate in order to avoid a narrowing of the problem per
spective and to take into consideration all relevant objections to the technology. It increased, however, the flood of information which had to be digested by the partici
pants. The subjects of the TA procedure, in fact, converged to a large degree with the questions which had to be examined anyhow for the registration of pesticides. This then raised the question what the genuine task of a TA should be—as this could not simply be a repetition of registration procedures we have already established.
The demands of some conference participants to make political issues, power struc
tures and action options in the field of conflict or the interests and strategies of main actors the subject-matter of expert reports were not met. Most of these questions have a normative touch and can only be dealt with to a limited degree empirically. For reasons of proportional representation, concurrent expert reports from the "other side"
would have been necessary. This would have "broken the bank". An expert report on ethical issues was commissioned, but it was expressly limited to the specific prob
lem of plant manipulation. The proposal to prepare an expert report which was to assess from the standpoint of ethics all the value judgements brought forward in the TA-procedure, with respect to HR technology, was rejected. In general, these deci
sions reflected the understanding that value judgements cannot and should not be delegated to experts. Moral and political evaluations are the domain of laymen (as citizens). In the TA procedure the competence for value judgements is with the par
ticipants themselves. These decisions were probably wholly justified. At the same time they meant that discussions of value problems were not automatically guaranteed and structured through the critical discussion of the respective expert reports. It remained a problem of the TA procedure how value discussions could be triggered and maintained.
5. Representative participation in the TA procedure
The question who should participate in the TA was, to a large extent, answered by the application for the procedure and by the decision on whom to invite to the first con
ference. Interest, political commitment and competence were the decisive selection criteria. The organizers and the WZB working group assumed that at least industry and the environmental groups, as the parties of social conflict, and the regulatory authorities should be involved. Furthermore, all areas of investigation had to be covered by experts from the relevant disciplines. This already led to such a high number of participants that the idea was dropped of also involving politicians from all factions and the customary representatives of public life (trade unions, churches etc.).
Representation of the media was guaranteed through invitations to the press.
As a rule, it was not difficult to assure the cooperation of scientists from the relevant research areas for the procedure by offering them expert report contracts. However, all other groups had rather mixed "feelings" about joining the procedure. The regula
tory authorities maintained a certain distance by seconding individuals who did not then appear as official representatives of their respective authority. Both German industry and the environmental groups had fundamental political reservations about participating in the procedure. This coalition against participation was certainly founded on opposing assessments of the situation. Industry was suspicious about being publicly "chastised". The environmental groups were apprehensive that their involvement would sap protest potential and would lend legitimation to technical innovations. Despite their ambivalences both sides decided to participate. The organ
izers tried to make it clear that the procedure was open in respect of its approach and its composition and did not favour any side. The question is whether that convinced the environmental groups in particular. What probably also played a role in overcom
ing resistance to participation is that it is not easy to publicly justify a refusal to par-
ticipate. In addition, anyone who refuses also runs the risk of the procedure taking place anyway—with the participation of the "other side" only.
At the first conference opposing parties criticised the composition of the participants as being imbalanced. In many cases, the calls for changes contradicted each other, e.g. both more scientists and fewer scientists were requested. Hence, there was little room for compromise. After all, there was apparently no real need for action. The divergent demands could be seen as an indication that those present represented the various positions in this controversial area. The Coordinating Committee once again prepared a proposal for solving the issue which was submitted to the participants after the first conference for comment. Silence was taken to mean approval. Taking up a demand from the participants an attempt was made to involve active farmers in the TA procedure. The attempt failed despite considerable efforts.
The number of participants in the procedure ranged from 50 to 60 (not including the WZB group). There were particularly high fluctuations amongst the representatives of associations and political groups. (The experts were contractually bound to regular attendance.) The Coordinating Committee supplemented the groups accordingly, without being forced to arithmetic precision, in such a way that the level of participa
tion remained more or less constant for the various groups.
If we break down the participants in the first conference according to institutions and groups, the following picture emerges (in brackets the breakdown prior to the final conference, see also appendix II):
Regulatory Authorities 1991: 7 (1993:10)
Research institutes (not including institutions associated with the environmental groups)
1991: 20 (1993: 25)
Environmental and consumer associations, alternative agricul
tural associations, associated research institutes
1991: 11 (1993:12)
Industry and trade associations 1991:10 (1993: 13)
What is more significant when assessing the representativity of the composition, is the attitude of the experts to the technology under review. Of the 42 experts and com
mentators, who had participated in at least one assessment conference, a slight major
ity was probably in favour of HR technology. According to an internal rating in the WZB group, 29-48% of the participants were in favour of the HR technology at the beginning of the procedure, 33-43% were against and 15-43% were neutral.10
6. Fair allocation of resources
The TA procedure constituted a major burden for all the participants, a burden which for many could not simply be tackled during their paid work. There were various kinds of problems with resources. Whereas time was probably the critical point for the participants from industry, participants from environmental groups, consumer organizations and the associated research bodies often faced financial difficulties when they were obliged to work in an honourary or self-employed capacity or financed through donations. In the procedure an attempt was made to compensate for this to a certain degree by ensuring that the experts from these groups were adequately con
sidered in the commissioning of expert reports and commentaries. This was necessary anyway for reasons of equal participation in the procedure.
What adequate consideration actually means may be a matter of argument. The fact is that in this TA procedure DM 124,000 (roughly 30%) of the expert fees went to sci
entists from the environmental groups or their associated research institutes. This sum does not reflect the overall volume of resources which were allocated to the critics of HR technology. If we include those university scientists who, according to the rating of the WZB group, were to be assigned to the group of critics at the beginning of the procedure, then this figures must be increased to DM 185,000 (45% of the total fees).
The question of distribution of resources was not problematic in the procedure up to the final conference either.
10 The major fluctuations in the rating within the WZB group do not correlate with the rater's own political preferences. They highlight the difficulty of the (external) assessment of attitudes towards HR technology. For an evaluation of self-assessment by the participants in the procedure, cf.
Bora/Döbert (1993).
7. Steering of the procedure: participants, Coordinating Committee, WZB working group
The central steering body for the procedure was the Coordinating Committee. This committee was set up at the first conference. In addition to the three organizers, it included three people from the side of the regulatory authorities, the environmental associations and industry. All decisions of the Coordinating Committee were submit
ted to all participants in the procedure for final comment. However, only rarely were there any reactions. Apart from a few exceptions, the decisions in the Coordinating Committee were approved unanimously, which made it difficult for other participants to break out of this consensus later. As a result the participants were more informed of the decisions of the Coordinating Committee rather than really called upon to decide on them.
The participants exerted major influence on the procedure by substantial contribu
tions, but not by formal steering. Only at the first conference did procedural matters play a central role but it was not possible to find a binding solution at that time. It is just not feasible that some 50 to 60 people with in some cases diametrically opposed interests are able to jointly conduct a TA procedure on the basis of self-steering—at least not within an acceptable period of time. More than 40 participants were already involved in shaping the contents by way of the expert reports and commentaries commissioned. Moreover, all the other participants were also entitled to submit comments, criticisms and evaluations in the form of papers. These were then to be discussed at the conferences and published in the documentation on the procedure.
The organizers themselves used this right in the final phase of the procedure by submitting "provisional conclusions" on the various problems in order to focus the assessment discussion at the final conference (see below).
The WZB working group had not been formally given the remit of steering the TA procedure but it did have real influence. All organisational work was undertaken by this group from the recruiting of the participants over the commissioning of expert reports to the planning and assessment of the conferences. For a period of more than two years, on average three people were fully occupied with this organizational work.
This extended greatly into the structuring of the contents of the discussions on the procedure. In formal terms, the WZB working group acted as the planning staff who continued to be dependent on the Coordinating Committee and who prepared and exe
cuted the latter's decisions. But this cannot conceal the fact that an "apparatus" of this kind can exert considerable influence. A certain degree of internal control was
ensured by the fact that the WZB group, itself, had been put together to reflect the various disciplines and political orientations pertinent to the controversial technology.
All strategic decisions in the course of the TA procedure were unanimously approved in the Coordinating Committee. This was the case for:
• the composition of the participants,
• the fixing of the subjects for the expert reports and the subsequent commissioning of the reports and the comments requested thereon,
• the form of publication of the materials and the results of the TA procedure, including the depiction of the various positions of the participants and consideration of these positions in the final synthesis report of the organizers (see below),
• the sequence of colloquia and assessment conferences,11
• the way in which the results of the TA procedure so far should be presented for discussion in the final conference.
In the Coordinating Committee (contrary to the situation in the plenary assembly of the participants) conflicts were settled through negotiation. All members of the Coordinating Committee had a factual right of veto. The remit for the steering of the procedure was subject to the reservation that the decisions at the "round table" had to be taken jointly. This did not exclude situations in which people let themselves be outvoted.
8. Loyalty vis-ä-vis the procedure
The TA procedure could only function because it succeeded in stabilising cooperative structures which served as the foundation for special loyalties amongst the participants and vis-ä-vis the procedure. On the other hand, given the divergent interests among the participants, this could only work because no excessively high loyalty was demanded. This applies in particular to the special behaviour which may be required of participants by the TA procedure for their actions outside that procedure.
11 Mr. Kiper, the representative of the environmental groups, was against the postponement of the final conference from February 1993 until June 1993. He felt that this could mean that the political character of the TA would be lost because of the extended duration of the procedure. The majority, however, felt that more extensive preparation of the contents for the final conference was absolutely essential.
The Coordinating Committee faced this problem on two occasions. The first was triggered by rumours that one of the participating industrial firms was going to submit an application to undertake field tests involving genetically modified sugar beet with HR as the marker. These plans were viewed as being incompatible with the TA pro
cedure and criticised as a violation of the behaviour expected in connection with the procedure. A continuation of the procedure was considered to be senseless if the industry involved were to introduce the technology without waiting for the results of the TA. This criticism was countered with the argument that a TA does not imply a moratorium.
No company could participate in a procedure of this kind if a condition were to be that it would renounce for several years the further development of the technology which was the subject-matter of the TA. That is undoubtedly correct. At the same time nothing would have prevented the Coordinating Committee from at least intimat
ing to the firm that applications for release during the term of the TA were not wel
come and would impose a strain on the procedure. A proposal of this kind from a few committee members was blocked by veto. Economic freedom of action was the essential precondition for the participation of industry in a TA procedure. The Coor
dinating Committee then simply decided to ask the firm for clarification of its plans.
A similar problem cropped up one year later with reversed roles. The environmental associations involved in the procedure presented their version of the results of the TA procedure to the press before the procedure had actually come to an end. In this con
nection, the Coordinating Committee remarked "that it neither wishes nor is it empowered to control public statements on HR technology by participants in the pro
cedure".12 The TA procedure is a limited area of cooperation, the public conse
quences of which are unforeseeable and do not have binding force for the political system. Under these circumstances it cannot legitimately be expected that those involved make major allowances for the TA procedure in their overall actions. Cor
respondingly, it was neither required that industry dropped the further development of HR technology for the duration of the procedure nor that the environmental groups stopped their public criticism of HR technology.
12 Minutes of the meeting of the Coordinating Committee, 5 February 1993. The majority of the
c o m m ittee merely criticised the fact that the press statement created the impression that the proce
dure was terminated.
9. Issues considered: risks, advantages, alternatives to HR technology
The contents of the TA procedure were largely determined by the subjects of the expert reports. There were two preliminary criteria for the selection of subjects:
(1) the arguments advanced in public for and against HR technology had to be dealt with;
(2) the TA should address the possible consequences of HR technology not the pos
sible options for weed control in agriculture. (The procedure represents what is termed a technology-induced TA not a problem-induced TA).
These preliminary criteria meant that the possible risks of HR technology became the central theme of the procedure. Discussion of the possible advantages was of far less importance and the alternatives to HR technology played only a limited role.
Of the twenty expert reports commissioned for the procedure (see appendix I), six dealt more or less exclusively with the risks of genetically modified plants. The expert report of the Öko-Institut (Ecological Institute) was a kind of reference report for the subjects covered. It summarised all relevant risk arguments. The other expert reports were related either directly as commenting reports or indirectly by way of parallel questions to that report. The paramount issues that emerged from the assessment discussion were whether the risks of transgenic plants differed from the risks of new plants which have been altered through conventional breeding techniques and how the hypothesis that there are special risks for genetically modified plants could possibly be justified.
Seven further reports addressed the possible toxicological and ecological risks of non- selective herbicides. Here, too, risk comparisons were to the fore in discussions.
Attention focused on whether the use of selective herbicides would worsen or possibly improve the situation vis-ä-vis the status quo.
The predominance of the risk aspect in the TA procedure had a whole series of conse
quences:
Firstly, it turned the TA procedure into a kind of prior examination for registration.
Sometimes it was difficult to comprehend why the problems had to be settled in a TA and not left to the competent authorities. This applied in particular to the potential risks of non-selective herbicides. In the case of transgenic plants there were no estab
lished registration tests.
Issues of judgement and valuation were given less emphasis than empirical aspects.
In most cases the factual preconditions of risk claims were controversial, i.e. the cau
salities, the probability and the scale of potential damage. The normative reference points of risk assessments, by contrast, were often uncontroversial. When toxic plant ingredients develop, registration will be denied.
The TA implicitly accepted the matter-of-course development of technology as the starting point. The hidden premise of risk discussions is a distribution of the burden of proof which favours technology: the banning of a technology requires justification, not the development and introduction of a technology. For much of the time, the TA discussions followed the pattern that the critics of HR technology had to make claims in relation to the risks of the technology and, faced with objections, were forced to substantiate their claims.
The predominance of risk subjects in TA procedures seems unavoidable. In a society in which technical innovation is built into the structure of science and industry and anchored in the constitution by a guarantee of individual rights, opposition to innova
tion is bound to resort to risk arguments. Not surprisingly, therefore, the main criti
cisms of genetic engineering are related to the possible risks. It is not argued that an acceptable need is not discernible or that technical innovation must in principle be subject to democratic parliamentary reservation. And, as long as risks are a central theme in the criticism of a technology, they must be the main subject of a TA. The minimum requirement of a TA is that it examines the legitimacy of the objections which are raised in public in connection with a technology.
Benefit analyses in the TA procedure concentrated on the claims of the advocates of HR technology that non-selective herbicides implied clear ecological and farming advantages. Here, different assessment frameworks were taken as the basis. On the one hand, even small ecological and economic improvements were chalked up as pro
gress vis-ä-vis the status quo. On the other hand, it was criticised that the improve
ments claimed did nothing to alter the fundamental problems of intensive farming.
They were simply variations of a trend which had to be seen as mistaken development and did not, therefore, constitute any benefit.
In claims to benefits the advocates of HR technology have the burden of proof. How
ever, this does not establish any symmetry to the burden of proof required of the crit
ics in respect of the risks of the technology. If critics fail to provide evidence of
relevant risks, the technology cannot be banned. If the advocates fail to provide evi
dence of relevant benefits, the technology cannot, nevertheless, be banned. At best, public support of the technology can be withdrawn. What "society" needs is normally decided on the markets. It is not the subject of political regulation. This regulatory structure took the sting out of the debate whether we actually need HR technology.
This is not necessarily so. One of the advantages of TA is that it can investigate more issues than can be politically regulated. Whether there is an acceptable demand for HR technology is a legitimate issue even when the answer can only have impact on public awareness and consciousness and cannot be directly translated into state regu
lation. Furthermore, it is also a legitimate subject whether socio-economic need should not be made the precondition for the registration of a new technology; even though this would mean a restructuring of the present legal constitution.13
Within the framework of the HR procedure this level of fundamental constitutional questions was scarcely touched on. On the one hand, routine reference was made to the demand on the market as a major indicator for social needs—which of course brought with it the question whether in this case labelling of genetically engineered products would not then be a minimum requirement. On the other hand (probably because of the missing regulatory relevance) the need question was not dealt with independently of risk issues. The thrust of the arguments was: the uncertainties of a new technology may not be offloaded onto society unless there is a real need for that technology. To that extent, the discussion of benefits brings us back to the discussion of risks.
Discussions of alternatives to HR technology played a marginal role in the TA. They were restricted to the level of technical details. Comparison with the potential and problems of mechanical weed control was a major point. The emphasis was, how
ever, on the main agricultural and farming framework conditions of intensive farming as the reference system. The alternative to this kind of farming, ecological farming, was covered in an expert report on non-chemical methods of weed control in the pro
cedure; it was not, however, the subject-matter of the investigation. All the same, repeated reference was made to the (existing or missing) performance, risks and costs of ecological farming. But there was no in-depth, factual clarification of these issues.
For this, it would have been necessary to have a parallel TA for ecological farming.
I3 See van den Daele (1994).
10. Evaluation conferences: science-based discussions and process orientation For the evaluation of the expert reports prepared in the TA procedure, discussion amongst participants was the decisive "arena". Discussions were held at two (two- day) conferences, by way of preparation three (one-day) colloquia were held on:
evolutionary biology (risks of genetic engineering), on policy conditions and targets for agriculture and on regulation problems with respect to HR technology. All par
ticipants had previously received the expert reports and a short five-page summary.
(See appendix III for the procedural steps of the technology assessment.)
When discussing the reports the experts for that respective area assumed a key role.
Participants who were experts in other subjects were edged out into an observer role or restricted to putting questions of a factual nature. In order to guarantee the active involvement of as many participants as possible and to foster controversial discussion, 15 additional commentaries were commissioned which were to introduce discussion on the respective subjects of the expert reports. From time to time the circle of par
ticipants was supplemented by suitable representatives of the "other side". The aim of the commentaries was to illustrate the various approaches and the various positions covered in the problem area and to stress the points which were clearly of importance for the political evaluation and regulation of HR technology. More particularly, they were to comment on:
• whether the central findings of the expert report were substantiated,
• whether important aspects were not covered by the expert report,
• which implicit and unexpected preconditions the expert report laid down,
• whether the conclusions were plausible.
The members of the Coordinating Committee took turns acting as moderator at the conferences. They were not responsible for summarizing the findings of the discus
sions but rather for ensuring that no points or contributions were omitted. It was their task, from time to time, to pool the arguments in order to structure discussion. The WZB working group was to support the Coordinating Committee in this respect.
The evaluation discussions presumably suffered less than is normally the case at con
ferences from time pressure. But they had two special features which were also con
straints: the discussions were scientific rather than political and they were process- rather than result-oriented.
The relative "reticence" of participants in respect of normative and political questions could be attributed to a certain pre-adjustment to the procedural rules. The plan was for problems of legal and political evaluation of HR technology to be dealt with as a priority subject at the final fourth procedural conference. It was only at that confer
ence that the legal expert report was dealt with which took up a series of legal policy questions concerning the regulation of HR technology. On the other hand, the inten
tion was not to postpone political questions to the last conference. It was far more the case that the Coordinating Committee had expressly indicated that the expert reports and commentaries should contain conclusions based on value judgements which should be seen as the starting point for political discussions within the TA procedure.
This, however, was only done to a limited degree. With the exception of the expert report on the ethics of plant manipulation, which was explicitly oriented towards a value judgement, all the other expert reports and commentaries mainly contained empirical and scientific arguments. One explanation could be the obligation of neu
trality attached to the expert role. What had more of an influence was probably the fact that the participants, themselves, defined their controversies as debates about empirical findings and not about values and goals. What was to be classified as dam
age was not a subject of controversy, what was controversial was whether the damage was possible or probable.
In the TA procedure no-one advanced the argument that the problems were not empirical but political in nature because it was a question only of what one wanted and what one considered to be acceptable. This would have meant, among other things, that the argument of the special risks of HR technology would at least have been qualified. Adherence to the relevance of the empirical questions was for various reasons. First of all, there was, more or less, agreement on many normative ques
tions which meant that new arguments on the valuation level could not be introduced.
Furthermore, the political clout of purely normative arguments is seen as being rather limited. In our society value issues (beyond non-controversial moral judgements) can be attributed to private pluralism or publicly answered by majority decisions. As a strategy of political critique it is, therefore, often more effective to attack factual claims which are based on valuations rather than to call for alternative values. At the conferences on the evaluation of the expert reports lengthy political discussions were the exception.
The evaluation discussions worked through central claims in the expert reports, i.e.
substantiation and proof were called for, objections raised, counter arguments formu
lated and examined. Often the sequence of arguments (in line with the sequence of requests to take the floor) went to-and-fro or in circles. However, as far as the con
tents are concerned, there was a factual accumulation of arguments. But in the proce
dure this accumulation rarely led to a declared convergency of positions along the lines of dispute settlement. In individual cases arguments were explicitly withdrawn and replaced by arguments on a different level. Sometimes there was a call for the Plenary to record the result of discussions as a declared consensus or dissent. This demand was not taken up. The moderators intimated that it would not be possible to sum-up ad hoc in a reliable manner the non-transparent state of discussion. The dynamics of the evaluation conferences were, therefore, process- rather than result- oriented. The discussions were consultations without conclusions or, if you like: the taking of evidence without an ensuing decision.
To a certain extent it seems appropriate that discussions in the procedure were not primarily designed to record consensus or dissent between the respective opposing sides. The negotiations on the expert reports were not a form of bargaining by means of which the parties could test whether a balance of interests could be achieved through give and take. Bargaining would, for example, have been involved if indus
try and the environmental groups had got together to agree on which policy they both could uphold for dealing with HR technology.14 The negotiations in our procedure could indeed be equated with cognitive debates on the interpretation of reality: along the lines of the taking of evidence. In these negotiations, however, the opposing sides do not have full sovereignty over the findings. Whether an assertion of facts is correct, whether a causal relation exists, whether a. hypothesis is plausible is not decided by the opposing sides alone. Consensus between the parties leads neither to the validity of empirical claims nor does this validity exclude dissent. In "contractual bargaining" the parties have the final say. The determining factor is whether they can agree or not. In the case of the "taking of legal evidence" it is the judge who has the final say.
Of course, this immediately raises the question of who can assume the role of the judge in the TA procedure. The procedure has been set up as a discourse with no third instance. Therefore it is up to the participants themselves to judge. If they can
not agree, then, by definition, the discourse has failed to produce results. Thus, the
14 Such bargaining may sometimes be involved in policy dialogues or consensus conferences; see Hingel (1993).
participants have not only process control, but also decision control.15 The conflict
ing parties may use decision control to prevent the taking of evidence from leading to judgements within the procedure. Decision control does not entitle them, however, to act as judges on their own behalf. If the participants cannot agree on empirical mat
ters within the procedure, they simply leave the judgement to observers o f the proce
dure. Normally it is the competence of institutionalized science to judge empirical questions. If the competence of science is challenged, the dispute can only be pre
sented to the public.
It is true nevertheless that the procedure changes its character if results of the TA can indeed only be presented from an observer perspective. What used to be a discourse with decision control by the participants becomes a procedure that transfers final judgement to a third instance. Decision making by third instances remains outside the rules of discourse. Correspondingly, it cannot lay claim to the specific legitimacy and integrative power of discursive conflict resolution.
11. Discursive procedure, pressure for consensus and statements of dissent
The final conference of the procedure had two functions. It was to bring to the fore the political problems of HR technology which had not been discussed so much up to then. It was also to do what the evaluation conferences had not done, namely to draw conclusions about the factual findings of the TA studies.
The evaluation and assessment of actions are viewed as political problems. For instance, should HR technology be registered, banned, promoted? Are the risks of transgenic plants acceptable? Where does the burden of proof lie for unknown risks?
Should new technologies be regulated in a different manner from old ones? Is an improvement in chemical weed control an advantage? etc. In the design of the TA procedure these questions were separated from information issues and subordinated to them in temporal and logical terms.
This sequence could scarcely be avoided given that TA is the examination and evalu
ation of the possible consequences of a technology. Before we ask whether a risk is acceptable we have to ask whether the risk exists. The question, for instance, whether we should accept the risk that transgenic HR plants might grow wild and
15 A distinction applied in procedural justice research; see Lind/Tylor (1988). According to this research, preferences for procedures seem to be more influenced by process-related variables (process control/voice) than by result-related variables (decision control).
establish themselves in natural eco-systems can only be dealt with appropriately if, prior to this, we have determined whether this process is at all possible, how likely it is and what the worst possible consequences could be. In the procedure, therefore, an attempt had first to be made to formulate the results of the "taking of evidence" which had become bogged down in the evaluation conferences. Of course, here, it could be that it is not possible to clarify the facts and that answers to empirical questions remain controversial. But this finding, too, had first to be recorded before any con
clusions could be drawn.
The WZB working group was assigned the task, by way of preparation for the final conference, of summarizing the findings of the discussions so far of the issues con
sidered. The summary had to be complete and clearly document the various positions of those concerned.
There were arguments in the working group about the method to be used. It was gen
erally agreed that first of all an overview should be established of all the arguments advanced in the procedure (in the expert reports, in the commentaries and discus
sions). To this end, argument trees were developed for the various areas in the pro
cedure in which, for each question, statements for and against were set against each other in blocks. What was controversial was whether this was to be carried further and the arguments classified according to their contents, i.e. reconstructed as a sequence of claims, counter claims, substantiation, objections, counter objections, etc.
The argument advanced against reconstruction of this kind was that it would contain hidden assessments. It would introduce into the statements of the participants nuances which may not have been intended by them at all. The argument advanced in favour of a reconstruction of this kind was that only this would truly reflect the course of the procedure. The block apposition of arguments would merely stress the divergent nature of the positions. The participants had, however, also produced convergence between these positions.
In fact the participants did not merely express and clarify their opinions. They con
tinuously argued with each other (i.e. against each other). The procedure prompted discursive communication: in the actual dialogue substantiation was requested and provided, objections raised and countered, evidence offered and examined. Hence, communication in the procedure differed greatly from the normal discussions in public political circles, which frequently do not move beyond the repeated exchange of announcements of positions. When people engage in discursive negotiations then, to a certain extent, they must open their positions to debate. In discourse no one has the
