Dr. Kerstin Cuhls and Dr. Hariolf Grupp
Fraunhofer Institute for Systems and Innovation Research
Breslauer Str. 48
D 76 139 Karlsruhe
firstname.lastname@example.org tel: *49 (721) 6809-156
fax: *49 (721) 6809-176
Foresight processes have been further developed during the past ten years, using different approaches. They are not based on the notion that the future is predictable, but that one can look into the future and try to shape the predictable paths. This contribution describes the development of the new foresight approaches in the nineties in Germany. New concepts evolved with a stronger focus on the communication about the future, the motives of policy-makers to conduct foresight and first effects of the new activities. Participation of the general public, a broader definition of "ex-pertise" and the notion of networking inside the different communities, bringing together interdisciplinary knowledge, are the major goals of the new foresight proc-esess. In this context, the German "FUTUR" is explained.
1 New Foresight Approaches
The end of the 20th century has witnessed the advent of many new foresight meth-ods and combinations thereof. Most of the experiences in organised experiments applying various foresight initiatives concerning future issues in science, technol-ogy or society were evaluated as very positive. Companies made use of the data, the media published a large number of articles, ministries reflected once more about their research priorities, and a research institution even based an evaluation on Del-phi results (Cuhls/ Blind/ Grupp 1998).
In most countries, the activities were supported by the research ministries or other public bodies. All foresight models try to implement communicative processes which integrate the different actors in the innovation systems. Most activities also attracted interest of the general public - either because of the approaching year 2000 - or because the need is felt to gain more information by looking into the fu-ture. The targets of foresight activities changed accordingly.
What is "foresight"? According to Ben Martin (1995a and b), foresight is "the systematic attempt to look into the longer-term future of science, technology, the economy and society, with the aim of identifying the areas of strategic research and the emerging of generic technologies likely to yield the greatest economic and social benefits". This definition has widened in the meantime to include not only technology foresight, but all areas of the human society.
Therefore, foresight is conducted in order to gain more information about things to come so that today's decisions are more solidly based on available expertise than before. Foresight is not equal to prognosis or prediction. Implicitly, it means taking an active role in shaping the future. A possible result may be that our expectations of today may be falsified in the future because of re-orientation. Former attempts to plan the future or to develop heuristic models of the future (in the sense of futurology) were based on the assumption that the future is pre-defined as a linear continuation of present trends (Linstone 1999, Steinmuller 1995, Flechtheim 1968, Helmer 1966). These approaches were not regarded as successful because they were too simplified. Some of them included different variables to match the complexity of the dynamics of the actual social, economic and technological developments, but this was also insufficient from the prognosis point of view. Nevertheless, some of these studies evoked a vivid discussion about the future (e.g. Meadows et. Al. 1972, Forrester 1971).
The new approaches in foresight can have many objectives (Cuhls 1998). In the context of policy-making, the most important are:
to get a larger choice of opportunities, to set priorities and to assess impacts and chances,
to prospect the impacts of current research and technology policy,
to find out new demand and new possibilities as well as new ideas,
to focus selectively on economic, technological, social and ecological areas as well as to start monitoring and detailed research in these fields,
the definition of desirable and undesirable futures and
starting and stimulating continuous discussion processes.
In reality, most areas of the future underlie conflicting influences, which at present cannot be assessed fully. Only some structural parts or framework conditions can be understood or partially influenced. If the knowledge in systems analysis theory is also taken into account, the mutual influences of systems and rules in which the action of mankind is embedded must also be reckoned with. An uncertainty was perceived in futures research when new experiences of the chaos theory emerged (Steinmuller 1995). The new thoughts that emerged in the nineties (starting with Irvine/ Martin 1984) did not state that the future cannot be influenced directly, but made clear that the influence on future developments is strictly limited and that the impacts can only partially be estimated. Nevertheless, the future can be "prospectively monitored". The accelerating changes that a person has to adapt to socially and psychologically make it necessary to anticipate these changes before they become reality (Helmer 1967).
During the last years, foresight studies were performed, especially in the USA, Japan, France, Italy and Germany, that deal with "critical" technologies (for an overview, see Grupp 1995 and 1999). They all followed different approaches and different quality parameters. Even the difinition of "critical" differs (Bimber/ Popper 1994). This kind of foresight study was mainly applied to set priorities. In the German case, one aim of the study on "Technology at the Beginning of the 21st Century" was also to provoke an intensive dialogue between six of the project agencies of the Federal German Ministry for education and Research.
Only a limited number of people were involved in these projects. This was criticised especially by the media and non-expert organisations, so that decision makers are now trying to initiate larger foresight programmes. Currently, the United Kingdom is one of the countries that have experiences in larger scale foresight activities. Similar approaches, but with fewer participants, can be found in South Africa (1997pp) or Hungary (1998pp). In these cases, the usefulness lies in a consensual priolity-setting and a communicative dialogue without the limits of science fields or disciplines. The only country with a longer continuous history of foresight up to now is Japan. In Japan, foresight is conducted on different levels, but there is not the one strategic approach (Cuhls 1998). Based on the experiences of the Science and Technology Agency, the first explorative attempts to apply the Delphi method here in Europe on a large scale took place in Germany. South Korea and France (and partly also Great Britain) loosely oriented themselves along these lines (Grupp 1999). These foresight projects were implemented in a more self-organised manner, e.g. by the use of foresight data for strategic planning of companies, setting priorities in research programmes or as basis data for the evaluation of the Fraunhofer Society (for an overview see Cuhls/ Grupp/ Blind 1998). A more limited approach entails less financial risk and might therefore be adequate for smaller countries.
Many foresight methods have already been developed to look into the future. They range from extrapolations via computer simulations to communicative procedures like future conferences, scenarios or the Delphi methods. Among the classical repertoire of methods, a method mix seems most promising. Not only quantitative but also qualitative methods with a strong communicative effect are regarded as useful and are becoming more and more "fashionable".
One of the major methodologies to structure a foresight discussion is the Delphi approach - as already mentioned above. "Delphi" as a method was initially developed by the RAND Corporation (Santa Monica, California) (Gordon and Helmer, 1964). The basic idea of a Delphi survey is to interview experts on a set of topics (in the case that will be presented here, dealing with major discoveries, technical innovations or large diffusion of technologies). The set of topics can be generated by the experts themselves or elsewhere. The aim is not only to collect the rough opinions of experts on certain future-oriented topics, but also to get each expert to react to the general opinion of his peers1. A set of questions is associated with each topics, such as the estimated degree of importance, the expected time of realization, the type of constraints, the need for international co-operation, etc. the procedure consists of a minimum of two stages. The results of the first phase (or round) are fed back to the experts during the second round, in order to provide them with an opportunity to change their original opinion or, on the contrary, to stick to it and explain why. In designing the inquiry this way, Olaf Helmar and his colleagues at the RAND Corporation intended to avoid uncontrolled psychological interference among the experts.
The method became popular when applied to a large-scale national technology forecast in the 1960s in the U.S.A. (Helmer, 1983; Gordon and Halmer, 1964). But if we observe the general use of the method until the present day, it can be seen that the diffusion of Delphi procedures world-wide has been largely concentrated in the area of strategic management of big firms and organisations. To illustrate the point: when Delphi questionnaires of the national inquiry were sent to experts from firms, public research organisations and administration in Germany, the firms' experts were revealed to be the least surprised and least critical about the procedure (but this dose not imply that they answered more readily, for strategic reasons). In the U.S.A., where the method was created, in more recent time we only know of smaller-scale studies were applied at firm or institute level after the first large public experience of the sixties.
Foresight, therefore, as a process may start with one method (e.g. the Delphi method): experts and persons from different disciplines or backgrounds are contacted and asked to think about the future. They also have to write down their thoughts and not only take this knowledge with them but diffuse it and provoke actions. The assumption is that synergy effects can occur, which means that the knowledge of several persons contains at least as much information as the knowledge of the single persons in sum and that this mutual knowledge exchange is fruitful. Group discussions can be especially helpful, because endorsing a group decision by yourself is more influential than just being convinced by another person.
Therefore, a mixture of methods like co-nomination (Nedeva et al. 1996), inter views, available address data bases, trade fair lists, publications, conference or workshop participation in different fields are necessary, as well as the mobilisation of people who are just interested (teachers, students, technology transfer personnel, etc.). it is especially necessary to involve the "non-experts" who are interested and motivated in bringing in their - often different - thoughts, especially about the application of the sophisticated ideas experts often have. Therefore, they must be placed in the position or situation to be or act as an expert, but also to make use of the tacit, implicit or hidden knowledge (Polanyi 1985) of people in general.
But who should be involved depends on the aim of the foresight activity. If it is a special field, mainly specialists are involved as "experts", as other stakeholders would not understand the matter sufficiently. If the foresight approach is broader, the general or interested public can also be involved. They might be able to judge the application of a development just as well as the experts although it is difficult for them to understand e.g. technical specifications.
The new foresight concept are designed rather as a facilitating factor for communication about the future than "finding out the truth about the future" - as concepts in the past often tried to do. Therefore, the definition of expertise is getting much broader. A major target is to involve enough persons, to make them communicate, find out if there is consensus or not, and create a certain commitment for activities.
The new German foresight activity is process-oriented. It takes into account the international experiences in the past ten years and makes use of a mix of methodologies. The process is called "FUTUR" (the Latin word for future), was started in 1999 and is planned to become a more integrated process that no longer separates the different foresight dimensions. Until now, in all countries, the studies centred mainly on scientific-technological questions and influenced the discussion in the other areas. Themes and topics in FUTUR shall therefore be broader and include dimensions like education, ethics, social questions, employment and education policy or resource allocation.
Differences in the various foresight approaches result also from the sponsorship and institution operating the studies or processes. In most cases, the research and technology ministry sponsors foresight activities within the framework of a "programme" and it only pays external institutions for selective additional studies. Ministries have more power and therefore better access to the different stakeholders in the innovation system. Therefore, the centrality in the organisation of a foresight process seems to be an advantage. But it can also be a disadvantage because of a lack of capacity and experience in the organisation of such large projects. But often "more neutrality" is asked for.
During the last years, the Delphi method and its variations gained some attention and were further developed in order to let the different actors in the innovation arena participate in looking into the future. All these participants are experts - although sometimes with a different point of view. And this is the starting point for the rather new experiments in some countries involving more actors than only "experts".
The new foresight processes also do not only look out for technologies and scientific developments (although these are major driving forces) but look at demographics, society, economics, the environment, politics, culture, and so on. The whole range of themes has widened, thus making foresight more complex and complicated. To introduce a structure into the puzzle, to keep the overview, to select the different parts and paths in more detail and then re-fit them into the whole is a difficult task which is often solved by using a mix of methodologies, e.g. scenarios in combination with surveys or Delphi studies.
2 Foresight Actors in Germany
In many countries, the existing methods are already very well developed, e.g. the scenarios in their different ways. Delphi studies or simulations are getting better and better with the increase of computer capacity, but the link to society is often missing. Other actors in innovation processes are much more used to applying data and information so that it is no problem for them to deal with foresight studies. Some examples for the use of the German Delphi '98 are:
Companies are using foresight data for their own strategic purposes (Cuhls/ Blind/ Grupp 1998). They analyse the framework conditions which are specific to them in a steadily changing environment, for example, a company-specific analysis can be made based on a set of Delphi results. The technologies or other developments that are relevant are selected, the data calculated and interpreted in the firm-specific context (Foresight forms a strong instrument together with additional information on detailed products or developments, on competitors or potential co-operation partners).
National foresight does not go into too much detail, but stops at the point where a marketable product is tangled. Intra-company foresight can fill this gap, but is not the task of the government. The framework conditions and broader contexts are given for small - and medium-sized enterprises (SMEs), at least. The problem here is that they often do not know that the large national studies exist or how to gain access to them, even if they do know about them, the capacity for own analysis or further firm-internal foresight is often missing. Nevertheless, foresight is also needed in SMEs.
Research institutions analysed their strengths and weaknesses based on foresight. Research has always to do with future(s), so basic research also has to be aware of what is going on in other laboratories and take all options into account. One example of a pre-evaluation is the Fraunhofer Society in 1998. During the evaluation of the system, an independent international committee was founded. The committee at first analysed the data of the Delphi '98 and developed a set of criteria which the Fraunhofer Society has to meet to fulfil their declared objectives, e.g. to do applied research relevant for the economy.
Then, a strength-weakness analysis followed, to find out in which fields the 47 institutes are active at present and if they are prepared to do research and development in the fields the Delphi '98 identified as relevant for the future (an index was calculated for this). With this approach, over- as well as under-capacities in certain fields could be identified. Conclusions were drawn on where new capacities are strategically necessary or where gaps can be left (Cuhls/ Blind/ Grupp 1998).
National, regional and communal policy can also make use of foresight approaches, be it as a strength- and weakness analysis like the one mentioned above for their research programmes, or be it for identifying priorities (like the companies do). Priority-setting is necessary because of shrinking budgets in every policy field. But policy also needs the commitment of all stakeholders in the fields. It can be very important for policy-making to find out if there is consensus or a conflict potential.
As the major sponsor of national foresight projects, the government also contributes by providing the data to all those who cannot do foresight studies on their own: small and medium-sized enterprises or the public.
A major problem for policy-makers is that success cannot be measured directly. There are no criteria to measure the success of foresight up to now. The evaluations made until now (like in the UK) are evaluation studies based on the question whether foresight meets the targets - but they cannot measure the dynamics which foresight processes set in motion. To count realised technologies (the forecasting aspect) is not feasible, because the not-realized projects are also a success if there was activity to promote or hinder these as a result of the foresight process (Kuusi 1999). The old example of De Jouvenel (1967) very well illustrates this: One foresees getting wet when it rains. This can be prevented by taking an umbrella. The forecast therefore is a failure - but for the person to do it and react on it, it is a sucess.
Media are also stakeholders in the process of doing foresight or just diffusing the information about it. Especially as a new century is approaching, the magic of the number 2000 makes the media aware of the future and future-related topics. The media have not been directly linked to or integrated in most of the foresight activities before now. That will change (see below). But journalists are independent. In democracies, they and their way of reporting on processes and facts cannot be directed by the government or national institutions. This means that they have the freedom to publish whatever they like.
In the first German foresight process of 1993, the media focused at first mainly on curiosities like the "baby-sitter robot". The journals criticised the Delphi approach in general by picking out the "typical Japanese visions". These visions had nothing to do with the approach itself and the "typical Japanese visions" were judged controversially by the German experts and sometimes even the Japanese participants (e.g. the 1,000m high skyscrapers or the artificial marine waste disposal sites). But after a while, the media also reported in a more differentiated manner: some Delphi topics were singled out and discussed that could be criticised, others were welcomed. The approach as such gained more and more attention, so that, when the report Delphi '98 was published, nearly all large newspapers and journals reported more neutrally. The media were a good mediator in diffusing the information that foresight data are available, so that all interested parties (from companies to students) got to know about Delphi '98.
The general public was only informed by the media or presentations about the project until now. Foresight as such remained the task of experts. But there were of course interested persons who asked for a better participation in foresight approaches and who are ready to contribute to the shaping of the future. Many of these people doubt that experts really know what "society" in fact needs or wants. In technology foresight, this means especially a shift from the "technology push" approaches to the more demand-driven "technology pull". But who represents "society"? the problem remains to define who is able to participate in such processes, who has enough knowledge to understand topics or questions and who is able to judge in a more objective way than just by subjective "emotions".
This leads immediately to the question: Who is an "expert"? Who can participate in foresight activities? Can "non-experts" also contribute to this discussion?
Many foresight studies up till now were based on Delphi studies or panels (expert groups). Both concepts involve only "experts". But looking at these studies in more detail, non-experts were also involved as the participants also answered questions in fields where they had little particular expertise.
In the Japanese and German Delphi studies, among others, for example the respon- dents were asked for a self-estimation of expertise, which was defined according to the following criteria (see Cuhls/ Kuwahara 1994):
high: This applies if the expert is actively working on the specific topic or equivalent topics in the field (including literature research). This is the own field of work.
medium: This should be chosen if the expert in the past worked on the theme or in the field, or when the expert is currently reading primary literature, but not working in this field anymore.
low: This should be chosen if the expert has read articles in newspapers or journals (secondary literature) or is in discussion with the appropriate experts in the field.
none: This means really no knowledge. Those who ticked this were not asked to answer this topic.
As the self-estimation was not regarded as sufficient, in the British Delphi study a co-nomination process was conducted to select the experts (Nedeva et al. 1996) by making the colleagues in the field recommend the experts. The problem here is that the scientific communities are very self-contained. Newcomers or "outsiders" with an uncommon opinion are rarely recommended. The idea of Kuusi (1999), to make detailed interviews with those who claim to be experts is a very interesting idea, but not practicable in large projects.
In the German Delphi studies, the major source of addresses were public databases, publications trade fairs, and associations, plus many other different recommendations. Sets of related criteria were worked out (e.g. belonging to a certain field of science and technology, regarded to be an expert, involved in research and development or something similar, the younger the better, to increase the number of female persons in the study). By telephone it was checked if the persons would really meet the criteria.
But this is not enough to secure the good quality of the sample. Also, the carefully selected experts cannot be "high level" experts for a whole technological field. In the Japanese-German comparison, it could also be found that the Japanese experts systematically claimed a higher expertise than the Germans. A reason for this can be a "better" pre-selection of experts in Japan. In Japan, in order to increase the response rate, a postcard was sent in advance to ask if the person would like to participate. But when some of the German answers were checked in detail and by name, it could be observed that German experts systematically understated their own knowledge because they regarded themselves as experts for the present, but doubted their expertise concerning "the future".
3 Networking and Participation in Foresight
In the German foresight activities, the need is now felt to involve more laypersons. The participation of all stakeholders in the innovation system is regarded as necessary to find out if there is consensus or conflict in the different future fields, because the different actors regard different aspects as more important than others or look at the matter from different angles. As one objective of foresight is the combination of different kinds of knowledge by the co-operation of different actors, it was also desired to install open Internet pages with the possibility of giving comments.
Additionally, the "objective" estimations from scientific studies can be seen more critically. Subjective career expectations of the different groups of scientists play a very large role - especially for innovations - and should therefore be controlled for in the assessments.
Participation is also relevant for the understanding of the results of a foresight activity. All participants should find their views represented somehow. This creates the expectation that "the rationality as well as the legitimacy of political decisions can be improved" (Hennen 1999, S.566). It creates a different stimulus to public discussions, especially if these are conducted transparently (in the case of FUTUR, by publication of the material, possibilities of participation via the Internet, a journal and so on), so that a sensitisation of the different actors and the public concerning technological societal, economic, political and other impacts may follow. In this way, foresight processes can also detect in which fields consensus already does or does not exist, so that in the second case a conflict potential can be identified in advance.
These kinds of learning processes were already designed in experiment, e.g. in discursive methods. Discursive approaches make for more rationalised discussions, because they focus on the need to provide arguments. They introduce reasons as a standard of political discussion. Therefore, they correct the strategic (party) intellectuality and argumentative propaganda which is common in the public (mass media) confrontations (van den Daele 1996, 129; also van den Daele 1994, 111ff, and van den Daele 1997). As discursive learning processes, mediation methods are also used in environmental policy (see e.g. WZB-Mitteilungen 1997; examples and conceptional discussions also in Koeberle/ Gloede/ Hennen 1997). Discursive approaches should therefore be examined by foresight teams with the different methodologies.
These new approaches are also trying to put the general public in the situation of being an expert; consumers especially are important to evaluate innovations. This takes into account that companies listen much more to the demands of their customers, because they sometimes have other wishes than those developed for them by experts in the labs.
The problem here is still the selection of those privileged to participate. In the German FUTUR, special attention is given to all kinds of teachers (in different types of schools and different faculties), for an optimal multiplier effect. As foresight will be relevant for future generations, students in particular will also be asked to participate. A certain knowledge level will nevertheless be a pre-condition.
In a foresight process, the government, or the Ministry for Education and Research (BMBF) in the German case, can implement a policy which relieves the government by involving many actors as well as promoting an "active society" in which the importance of a systemic interaction between the relevant actor groups is increasing, as a reaction to deepening interdependent relationships inside the policy arenas (health policy, higher education policy, industrial sector policy) (Messner 1995). Already Etzioni (1969) describes that collective knowledge and therefore the efficient performance of all actors in society and their capability to exchange information result in a steering resource similar to power or money. The state gains power by better legitimations of their decisions following its initiatives for facilitating debates about potential futures.
According to its definition, foresight is oriented towards the decision to go in certain directions. Therefore, it is a special challenge to co-ordinate the different actors in such a way that both, policy-makers and socio-economic actors, are able to perceive an advantage. One of the objectives of foresight is therefore to build upon existing or create new networks as lively units and to activate them. These networks cannot be steered directly, but can only be mediated indirectly. One of the connotations here is "Distributed Intelligence" (see picture 1)(Kuhlmann et al. 1999): Different actors in the networks and the arenas of research and education policy make use of the more or less "intelligent" instruments and information for their strategic future planning. The actor groups have very different value orientations and images of the future. These have to be interrelated in order to facilitate joint strategic use, but not to bring them into the same structure. This can be organised on a joint platform, on which controversial disputes as well as consensus can be sought, but on which still the different roots of the concepts remain visible.
It is very difficult to bring the different actors in the foresight process together. One reason are the differing interests, the others are language and information barriers (e.g. "the public" often does not understand the specific vocabulary of scientists, or scientists from different disciplines do not understand each other). The identification of the actors (from public data bases to member lists of associations, communities, schools) needs very different approaches, which should be followed by neutral parties so that a certain balance is possible. For this, selection criteria should be worked out (e.g. age, expert knowledge, field of expertise, sex, interdisciplinary interests), because in the new processes, not only technical-scientific but also social, societal-cultural, economic, educational and other questions are raised.
Identifying experts and other participants is the one task, to link them and make them communicate is the other. Therefore, an interactive network has to be built up. Following the idea of the network, the representatives of the different actor groups will already work out questions in a very detailed manner in different teams, but on the other hand also contact each other independently, without contacting the whole group formally.
Therefore, it is first necessary to find out who the actors in the different fields are (e.g. in FUTUR in "Mobility and Communication" or "Health and Quality of Life"). Similar to the co-nomination process, every actor can suggest other persons, so that a dynamic and comprehensive network can be rapidly created. But care must be taken to include also those actors who cannot be found by this method, e.g. outsiders of the communities, persons working along interdisciplinary lines, students, the disadvantaged or those who for other reasons are just not "community members". This can be done to a certain extent by statistical approaches in data bases, or by other research with interviews, but it must be done as neutrally as possible. It should be tested, if co-nomination can also be used to identify "the least acknowledged person" in the field.
One of the objectives of the German foresight process FUTUR is the openness for other topics and themes, plus the participation of non-experts. The interpretation of already existing data and the creation of new questions cannot be independent of different opinions on problems, especially for the assessment of new technology and its application effects. Therefore, the identification of future topics and the selection of those which deserved in-depth treatment needs the participation of those involved and interest groups or lobbies. Furthermore, the specific knowledge of the different actor groups shall be applied in order to detect risks and chances, but also to arrive at innovative solutions for existing problems. The representation of as many opinions or sides of the problem as possible from the different actor groups in the various fields can improve the analytical depth of the assessments and the commitment for a topic.
4 Organisation of FUTUR- The New German Foresight Process
In the German FUTUR process, the foresight activities include the collection, the bundling, and the analysis of themes, as well as the diffusion of the results. The activities are supported by 1. selected prospective teams as representatives of different interests, 2. the media, and 3. the Internet as an interactive tool or platform for the discussion. There are also certain milestones of the process which should be followed by evaluations of neutral, if possible, international partners. The distinction between a Pre- or Post-Foresight-Phase is no longer possible. What is more important is the continuity of the whole foresight process. As the experts stem from different fields, the approaches have been broadened and the experiences applied to other fields.
At first, topics and questions concerning the future have to be raised. The different kinds of actors will work on these. In this first phase, also new actors in the fields are identified so that a network of them ("intelligence pool") can be created. The groups meet physically in workshops or virtually via the Internet. The task is to work out questions in discursive processes supported by creativity and other techniques (e.g. scenarios). These questions must be estimated as worthy of discussion on a broader basis. Also mediated future conferences can be conducted in this phase. The criteria for this work are given for example in the form of political questions from the ministry itself. Possible questions are: Which topics increase the competitiveness of a country? Which fields need kinds of education that were not provided before now? How can this be organised? In which fields does a conflict potential exist? How can the government avoid or prevent these conflicts?
After many questions and topics have been raised, a special focus is needed. This can be done with a comprehensive, systematic and comparative survey among the most important actor groups. This central study should include the objective estimation of technological potentials as well as the subjective wishes (also of the different parties involved), whose influence on future developments cannot be underestimated. The Delphi methodology with its feedback to all participants is one of the methods that can be used in this context. Scenario workshops with feedback could be another possibility.
A kind of ideal foresight process can be described recursively, as in picture 2: when the intermediate results of the structured surveys are available, the implementation and discussion follows. For this, teams from the different actor groups are founded in order to discuss the future topics in a target-oriented manner. These discussions follow applied discursive methodologies. So-called planning cells, consensus conferences, focus groups, mediation (especially in the case of conflicts) or local scenario workshops can be used to structure the discussion. These methodologies can only be applied if the topic itself is already clearly stated. The conception, mediation, and monitoring will be designed by a neutral partner. For in-depth discussion in specific fields, a specialist mediator is needed who is able to ask the "right" questions and to follow the debate. This mediator needs to behave as neutrally as possible.
Especially, in the implementation phase, it must be determined exactly which targets are to be pursued and whose targets they are. For the BMBF, setting priorities in certain research programmes or working out new research programmes can be a target. Other targets can be more interesting for industry (e.g. tax reductions for certain fields, financing). Specialist evaluations the recommendations of the public ("citizens' will) for certain projects (e.g. Should the German magnetic leverage train "Transrapid" be built or not?) or definite plans for measures (financing of a specific research institution, the foundation of new institutions or closing down no longer needed institutions) can also be objectives.
The allotment of tasks in FUTUR is described in picture 3. Three major actors are involved in the management of the process. One part is to monitor the process, prepare the concepts and analysis, and organise the evaluation of the whole process.
The second major part is that of decision-maker and mediator of the whole process, to be played by the BMBF in general. The minister herself will take responsibility for this process. The various media and especially the Internet are the liaison partners in the process. A management and co-ordinating institution is also needed to manage the Internet, invite the different kinds of actors and co-ordinate the process. This ideal concept looks more static. As a network, it will look like picture 4.
All actors in the fields and the organisers are involved in a large network and cooperate very closely. The Internet is the platform for transparency and for presenting results and intermediate results. The external presentation part will be a large role because without activating and motivating people (also the public) to participate, it will be difficult to keep the process alive. In the Internet, information will be provided on already existing material as well as links to other homepages that are interesting in this context.
The prospective teams are kind of panels consisting of the different actors. They work dependently and independently - according to their wishes. They can be selforganised and are assisted by a scientific secretary. A person who is well known in the field and who is able to motivate the participants will be one of the key persons in the group. The teams work out their topics using the different foresight and creativity methods. They are assisted by professional mediators and are asked to deliver an outcome after a certain time. This kind of outcome is not fixed in its form. It can be a report, recommendations, short descriptions of new trends, CD ROMs, workshops, the organisation of a conference, founding of a specific network or task forces. But the outcome is necessary to get some visible return out of it, and it should also be of use for the participants themselves.
5 Outlook and Difficulties in Doing Foresight
The German foresight process FUTUR is just beginning, but already the difficulties of co-ordination are apparent. The network of actors and participants is already large and becoming increasingly larger. Also the organisation is co-ordinated in a team consisting of different institutions. Who has to play which part? The whole process management is supposed to be as self-learning as the process itself. It also needs time to convince all possible partners, to provide the financial support and solve the technical and organisational problems of an Internet page.
But the major difficulty is to convince actors in the innovation process that they really have an advantage from participating in foresight studies, i.e., investing their time in that process. If they are convinced, additional, equally competent participants have to be found and persons from very different fields integrated, not forgetting potential partners. In some fields, the interest is already very great - especially in industry. In other fields, this is not the case, so that it is difficult to involve the "right" mixture of participants.
The next challenge is to develop new methodologies for these new approaches. Some are already available, but the mix of methods can also be very useful, depending on the field. Scenarios can be conducted in many different ways. Should they always be applied? Are open panel sessions possible? What about the mediation or creativity techniques? Are they adequate? Can persons really be made aware of the future tasks instead of remaining in the thought mode of the present (which was often a problem in earlier foresight studies)? How can the tacit or implicit knowledge be activated?
Another open question is: How can the participants be permanently motivated? Everyone is very busy and has other tasks too. Is the Internet really accepted as the instrument for participation? Are workshops enough as an addition? Or do we need "real events"? We would recommend them, but which ones? Who should organise them?
And can the different targets of the different participants really be met? The expectations are very high. Often, the organisers evoked these high expectations. It is already difficult to explain that the concept aims at different targets for different actors. This makes it often very vague. Too vague?
What about the evaluation? The foresight processes have to be evaluated by outsiders (meaning foreigners, in the German case). How can this be organised? The evaluators should understand the German language in order to go into details and follow the discussions. And they should have enough knowledge on foresight to judge the matter. What are the targets that can be evaluated when every participant has his own reason to participate? What is the success of foresight? If it changes something? If it motivates the stakeholders in the system to act in a certain way? If foresight supports the communication processes? Or if co-operation partners find each other? How can a process like this be "measured"? And how can the outcome be translated into a certain "strategic thinking" (in the sense of Godet 1997)?
Many questions are still open, and this contribution was not supported to answer all of them. Its major aim was to inform the interested participants about what is going on in Germany as regards the new foresight approaches. It has also shown some of the limitations foresight processes have to face. But change is necessary to be better prepared for the tasks of the future. In Germany, we would comment this with "Only change is unchanging".
Bimber, B. and Popper, S. W. (1994):
What is a Critical Technology? , Paper of the RAND Corporation, Santa Monica, California
Bundesministerium fur Forschung und Technologie (BMFT, Ed.)(1993):
Deutscher Delphi-Bericht zur Entwicklung von Wissenschaft und Technik, Bonn
Cuhls, K. (1998):
Technikvorausschau in Japan. Ein Ruckblick auf 30 Jahre Delphi- Expertenbefragungen, Physica, Heidelberg
Cuhls, K.; Grupp, H. and Blind, K. (Eds.)(1998):
Delphi '98 - Neue Chancen durch strategische Vorausschau. Tagungsband der Tagung in der Deutschen Biblio- thek in Frankfurt/Main am 1. Juli 1998, Fraunhofer Institute for Systems and Innovation Research, Karlsruhe
Cuhls, K. and Kuwahara, T. (1994):
Outlook for Japanese and German Future Technology, Comparing Technology Forecast Surveys, Physica, Heidelberg
Edquist, Ch. (Ed.)(1997):
Systems of Innovation. Technologies, Institutions and Organizations, Pinter, London/Washington
Etzioni, A. (1968):
The Active Society, New York
Flechtheim, O.K. (1968):
Futurologie - Moglichkeiten und Grenzen, Frankfurt/M. / Berlin
Forrester, J. W. (1972):
Der teuflische Regelkreis. Das Globalmodell der Menschheitskrise, Stuttgart
Godet, M. (1997):
Scenarios and Strategies. A Toolbox for Problem Solving, Cahiers du LIPS, Special Issue, Paris
Gordon, T. J. und Helmer, O. (1964):
Report on a Long-Range Forecasting Study, Rand Corporation, Santa Monica/California
Grupp, H. (Ed.)(1999):
Technological Forecasting and Social Change, Special Issue on National Foresight Projects, Vol. 60, No. 1, Jan.1999, Elsevier Science, New York 1999
Grupp, H. (Ed.)(1995a):
Der Delphi-Report, dva, Stuttgart
Grupp, H. (Ed.)(1995b):
Technologie am Beginn des 21. Jahrhunderts, Schriftenreihe des Fraunhofer-Instituts fur Systemtechnik und Innovationsforschung, vol. 3, second ed., Heidelberg
Helmer, O. (1983):
Looking Forward. A Guide to Futures Research, Beverly Hills/ London/New Delhi
Helmer, O. (1967):
Analysis of the Future: The Delphi method, Rand Corporation, Santa Monica
Helmer, O. (1966):
Social Technology, Basic Books, New York/ London
Hennen, L. (1999):
Partizipation und Technikfolgenabschatzung, in: Brochler, S.; Simonis, G. and Sundermann, K. (Hg.): Handbuch Technikfolgenabschatzung, Vol.2, Ed. Sigma, Berlin
Irvine, J. and Martin, B. R. (1984):
Foresight in Science, Picking the Winners, London/ Dover
Jouvenel B. de (1967):
Die Kunst der Vorausschau/ L'Art de Conjecture, Neuwied/ Berlin
Koberle, S.; Gloede, F. and Hennen, L. (Hg.)(1997):
Diskursive Verstandigung? Mediation und Partizipation in Technik-Kontroversen, Nomos, Baden-Baden
Kuhlmann, S. et al. (1999):
Improving Distributed Intelligence in Complex Innovation Systems, Final Report of the Advanced Science & Technology Policy Planning Network (ASTPP), Karlsruhe
Kuusi, O. (1999):
Expertise in the Future Use of Generic Technologies. Epistemic and Methodological Considerations Concerning Delphi Studies, Helsinki School of Economics and Business Administration, Helsinki
Linstone, H. (1999):
TFSC: 1969-1999, Technological Forecasting and Social Change, Vol. 62 (1999) p.1-8
Martin, B. R. (1995a):
Foresight in Science and Technology in: Technology Analysis & Strategic Management, Vol. 7. (1995), No. 2, p. 139-168
Martin, B. R. (1995b):
Technology Foresight 6: A Review of Recent Overseas Programmes, HMSO, London
Meadows, D. L.; Meadows, D. H.; Zahn, E. and Milling, P. (1972):
The Limits To Growth, Universe Books, New York
Messner, D. (1995):
Die Netzwerkgesellschaft. Wirtschaftliche Entwicklung und internationale Wettbewerbsfahigkeit als Probleme gesellschaftlicher Steuerung, Weltforum Verlag, Koln
Nedeva, M.; Georghiou, L.; Loveridge, D. and Cameron, H.:
The use of conomination to identify expert participants for Technology Foresight, in: R&D Management, Vol. 26 (1996) No.2, pp. 155-168
Polanyi, M. (1985):
Implizites Wissen (The Tacit Knowledge), Frankfurt
Steinmuller, K. (1995):
Beitrage zu Grundfragen der Zukunftsforschung, WertstattBericht des Sekretariats fur Zukunftsforschung 2/95, Gelsenkirchen
van den Daele, W. (1997):
Risikodiskussionen am "Runden Tisch". Partizipative Technikfolgenabschatzung zu gentechnisch erzeugten herbizidresistenten Pflanzen, in: Martinsen, R. (Eds.): Politik und Biotechnologie. Die Zumutung der Zukunft, Nomos, Baden-Baden, p. 281-301
van den Daele, W. (1996):
Rationalitatsgewinn durch diskursive Verfahren?, in: Holzinger, K. and Weidner, H. (Eds.): Alternative Konfliktregelungsverfahren bei der Planung und Implementation großtechnischer Anlagen, WZB: Mimeo, Berlin
Van den Daele, W. (1994):
Technifolgenabschatzung als politisches Experiment. Diskursives Verfahren aur Technikfolgenabschatzung des Anbaus von Kulturpflanzen mit gentechnisch erzeugter Herbizidresistenz, in: Bechmann, G. and Petermann, Th. (Hg.): Interdisziplinare Technikforschung. Genese, Folgen, Diskurs, Campus, Frankfurt/ New York, p. 111-146
Abfall: Mediation in Berlin vorbereitet, in: WZB-Mitteilungen 76, Juni 1997, S. 19-22
A pragmatic methodological description of Delphi was published by the Federal Ministry of Research and Technology (BMFT, since 1994 with the acronym BMBF)(definition of the Fraunhofer Institute ISI, which was the operator of the German Delphi survey, see also Cuhls 1998): "A survey among experts in two or more rounds in which from the second or later rounds the results of the previous ones are included. Therefore the experts judge from the second round under the influence of the other experts' opinions. The Delphi method is a relatively strongly structured communication process in a group. During this process those facts are assessed, for which there is uncertainty and no full knowledge available yet."