Chapter 8

A method for co-operative analysis and definition of complex interdisciplinary societal problems

Based on the material described in the above chapters one to seven, we developed a method[1] for assisting the analysis and the defining of complex interdisciplinary societal problems. This method is an example of how the analysis of complex interdisciplinary societal problems can be supported. The method gives no algorithm for the solution of a problem, only guidelines, suggestions and heuristics for how to handle a problem. The method should be regarded as a framework in which steps for analyzing are given. At each step one or more methods and tools can be added. Basically the steps should be approached sequentially, and should be traversed, several times, iteratively.

Hardware tools : group support room[2], computers, flip over, overhead, projector, white-board, printer, external communication facilities[3], copy machine

Software tools : hypertext system, system dynamic simulation software, spreadsheet, database, text-writer, electronic mail[4].

Compram is an abbreviation of COMplex PRoblem Analyzing Method. The method COMPRAM, assists the process of co-operative problem analyzing and defining of complex interdisciplinary societal problems. The method can be considered as a special kind of group decision support method (GDSS). However it does not primarily assist in reaching the decision, but rather the communication, the exchange of information, and therefore we prefer to call it a group information exchange method (GIEM).

COMPRAM is a framework consisting of a combination of methods and tools. The method assists the first sub-cycle of the problem handling cycle[5], specifically the construction of the conceptual model of a problem. The conceptual model contains the description of the relevant concepts and the phenomena and relations between the concepts and the phenomena involved in the problem. By describing the conceptual model the problem is defined.

As we have already described in chapters three and seven, at the start of the process of problem definition each participant often has a detailed view of that part of the problem related to his or her expertise and a vague mental idea of the whole problem, which is colored by his or her domain view, political point of view, cultural background and temporarily situation. The mental ideas of the participants about the problem will therefore differ from each other. The detailed view of all the participants should be combined, as far as possible, in an overall conceptual model of the problem, following intensive information exchange, discussion and data gathering. The method is based on ideas coming from cognitive psychology, computer science, theories about group-processes and ideas of system dynamic modeling and social science in general[6].

Individual preparations are alternated with group sessions, in order to combine the benefits of the different ways of working. In the preparation session the facilitator and the participants work either individually or in small sub-groups. The benefit of the preparation sessions is that these can take place at any moment and place that suits the individual. Due to limited time and possibilities for the participants to come together, the method will stimulate individual preparations whenever this is possible. In the individual sessions the emphasis is on concentration, contemplation and thinking, in the group sessions the emphasis is on the interactions and mutual influencing of the team members. A third way of working is in small mono-disciplinary sub-groups consisting of the participant and his or her reference group, which consists of a group of colleagues of the same domain. This group consists of three to four people and functions as a knowledge support group. The reference group will not participate in the group sessions.

The method combines social scientific methods of information exchange with technical (computer) support. The data gathering will be based on discussions, (small) lectures, or papers of the participants by which they explain components concerning their field. This kind of data gathering will be combined, when necessary, with aspects of other methods, such as interviews, observations and literature search. The information exchange will be combined with specific social scientific methods for group information exchange such as brainstorming and ranking. The method is technically supported by overhead projectors, computer tools and communication facilities such as fax, telephone etc. Software tools such as Groupware[7], databases, text-writers and spreadsheets support the data gathering and data exchange. The problem handling process will be assisted by descriptions in words and graphical descriptions by hypertext, by system dynamic software and other modeling software, for instance COPE[8] for cognitive mapping. The software tools, especially the modeling tools, serve as a common language for the participants. The problem defining process be assisted by a facilitator and also by a person who handles the software, the technical assistant[9].

1 The participants have only partial, often very well detailed, views of the problem which differ from each other.

2 The partial views of the participants will influence the knowledge of other participants.

3 Even after an intensive information exchange there will be white and blind spots, and contradictions in the data and in the knowledge of the participants.

4 The method can only to a limited extent prevent the negative influences of power differences in the discussion such as happens in group think.

5 Hidden agendas can only to a certain extent be avoided or their influence neutralized.

7 Expressing a problem in different languages using, apart from 'natural' language, a graphic language, will extend the understanding of the problem.

8 By defining concepts and theoretical ideas carefully, the communication in a multi-disciplinary team will be facilitated.

9 The seven layer model[11] can serve as a common team language during the problem handling process.

10 For analyzing a complex interdisciplinary societal problem, considerable time is required, time to work together, to think and to discuss things repeatedly. Several (individual and group) sessions are therefore required.

There are numerous kinds of complex interdisciplinary societal problems[12]. The method, described here is primarily adapted to assist the handling of knowledge problems[13]. The method focuses especially on those knowledge problems which have to do with causal relations. The method is specially developed for stimulating the exchange of information in a multi-disciplinary team[14].

The method only supports the last three phases of the first sub-cycle of problem handling. These are the phases following awareness of the problem and the conviction that the problem can and should be handled. One will also know who the (temporary) initiator of the process is. The phases lead to construction of the conceptual model of the problem, and hence to a definition of the problem. The phase of awareness is excluded, because this phase differs from all the other phases in the problem handling cycle (Outshoorn, 1986).

The process of co-operative problem definition begins by selecting the participants. This will be based on the mental idea the initiator(s) have of the problem. Selecting the participants not only depends on the (major) domains that are involved, but also on the kind of problem and on how much is already known about the problem. When it becomes clearer in the problem handling process that other domains are involved the composition of the team can alter.

As indicated earlier, the selection of participants is very important. It influences the outcome of the process, it includes and excludes already certain 'solutions' and it has an effect on the credibility and the acceptance of the definition of the problem by others.

The problem handling process of the participants is guided by a facilitator who will not interfere with the content of the discussion. The facilitator is neutral and responsible for the group and the problem handling process.

The participants together are responsible for the outcome, with respect to the content of the process. In order to avoid group think several measures are taken, such as letting someone take the role of the devil's counsellor, asking contra-expertise and inviting persons from outside to join some of the group sessions and by conducting anonymous brainstorming and voting. Each participant is supported by a domain reference group, with which they can discuss the part of the problem concerning their domain field.

Constructing the conceptual model of the problem will be supported by a seven layer model[15]. The seven layer model is meant to express the problem in different ways, using different languages and different models which can help to give a better view of the problem.

The seven layers together represent, in the end when all the layers are sufficiently filled, the conceptual model of the problem. The content of the layers is:

III Verbal description of the theories, hypotheses, assumptions, experiences, and intuition, that explain the influence of the concepts and the phenomena on each other

V A semantic model which is a graphic representation of the relations between the concepts and the phenomena

VI A causal model which is a graphic representation of the causal relations between the concepts and the phenomena

VII A system dynamic simulation model[16]. The system dynamic model is in this sub-cycle used as a discussion vehicle.

The layers I to VI are supported by hypercard software. The layer VII is supported by system dynamic software.

In layer I the entire problem is described in words. A natural language, which each participant can understand, is used for this.

The concepts and the phenomena used in the description of the problem in layer I are defined in this layer. In this way the participants are stimulated to operationalize and define the concepts and phenomena they use. This gives other participants the opportunity to learn the concepts of other professions and it prevents verbalism. The concepts and the phenomena in layer II are connected by the hypercard software with the same concepts and phenomena mentioned in the description of layer I. The definitions in layer II can be reached by 'clicking' on the words in the description in layer I and vice versa. This layer is also connected to the other layers of the model.

In layer III the relation[17] between the concepts and the phenomena of the problem are described. These relations are based on theories, hypotheses, assumptions, experiences or intuition. The connections between the descriptions can be indicated by a character and a number[18].

This layer is connected in the same way as described above, with the description of the problem in layer I, with the definition of the concepts and the phenomena in layer II. It is also connected with layers IV, V and VI, and related to layer VII.

In layer IV a graphic representation of the knowledge in the knowledge islands of the problem is made.

The way the knowledge islands are filled indicates the completeness of the knowledge. This layer is connected with layers I, II, III, V and VI and related to layer VII.

In layer V a semantic model of the problem is made. A semantic model is a graphic representation that, in this case, indicates the relation between the concepts and the phenomena involved in the problem. In the semantic model the concepts and the phenomena and the relations can be indicated by the same characters and numbers as described in layer III. This layer is likewise connected with layers I, II, III, IV and VI, and related to layer VII.

In layer VI a graphic representation of the causal relations between the concepts and the phenomena is made. In the causal model the concepts and the phenomena and the relations are numbered the same way as in the other layers. This layer is connected with layers I, II, III, IV and V, and related to layer VII.

In layer VII a system dynamic model of the problem is made that is based on the causal model. This layer is connected to the other layers, although not directly by software[20].

Parts of the problem and the different domain knowledge can be worked out in more detail in sub-sheets of the layers I to VII. The sub-sheets of one domain are connected with each other and are connected with the overall problem. It is often necessary to work out a part of the problem in detail in order to get a better view or because otherwise the models are too large to oversee.

As said earlier, the process of problem definition alternates group meetings with individual preparation sessions. The number of group sessions depends on the complexity of the problem and the time available of the participants. The process of analysis is an iterative process of describing the problem, making an inventory of the concepts and the phenomena that are involved in the problem, defining the concepts, defining the phenomena, describing the theories, hypotheses and assumptions, experiences and intuition, drawing the knowledge islands, forming hypotheses, finding supporting data and data that is in contradiction with the theoretical ideas, constructing the semantic model and the causal model and the simulation model. Apart from the last part, this will also carried out for the domain knowledge of different domains. The semantic, the causal model and the simulation model can be used to explore how things are related. Together these descriptions form the conceptual model. The conceptual model consists of things that one knows and things which are assumed, or on which one has hypotheses about.

The first sub-cycle of problem handling is complete when the participants together agree that the problem is defined. Because, as we said before, many societal problems are imbedded in a dynamic context and the problem itself is constantly changing, this kind of problems only can be temporarily defined.

Analyzing a complex interdisciplinary societal problem this way is an intensive way of working, and takes a large amount of time, from the facilitator as well as from the participants and their reference groups. However analyzing complex interdisciplinary societal problems is difficult and should be taken very seriously because the interests of many people are involved.

In order to illustrate the way the method can be applied, we give an example. The procedure we are going to describe here is a general one. The actual problem or the circumstances may force one to alter or to adjust the procedure. We will only give a global description, describing only the major steps[21].

The facilitator should not follow these directions blindly, but should use them as guide-lines. It is very important not to disturb the group process, only to guide the process in the prescribed direction. The products and initial sequence however should be maintained. As explained earlier, the number of sessions depends on the problem, the time the participants have available, the kind of problem and on how rapidly the process proceeds.

The time between the sessions depends on the minimum preparation time that is needed. The participants should reserve, at least (half) a day a week for preparation. The time between the sessions can be shortened by asking the participants to reserve one or two days a week for analyzing the problem. In the group sessions Groupware and other software tools are used. The individual preparation sessions are supported by individual software tools such as text writers, databases etc. In group sessions data exchange of all kinds is central, in individual sessions the emphasis is on reflection, thinking and debating in small domain groups. All the sessions can be combined with moments of observation, literature study and other data gathering methods, either by the participants themselves or by a special research team. In the case of a very important problem, two teams can be asked to analyze the same problem in parallel. Near the end of the problem handling process the teams can compare their results and discuss their similarities and the differences.

As an experiment of thought, we will illustrate some of the products of the sessions with a real life example of a complex interdisciplinary societal problem. We have selected the Aids problem[22] for this purpose[23].

The Aids problem is a complex interdisciplinary societal problem because it affects large parts of the society and it involves many disciplines. The phenomena that influence or are influenced by the problem belong to the knowledge domains of medicine, psychology, sociology, education, economy, law, politics, theology etc[24].

In order to begin handling the Aids problem, one must be aware that there is a complex interdisciplinary societal problem called Aids, and that something can and has to be done about it. In order the handle the problem correctly, the problem should first be defined.

- to have a first global description of the whole problem, including what domains are involved, which concepts and which phenomena

Forming a mental idea by hearing, thinking, reading, talking, writing and asking questions about the problem

P1-2 General preparation: a brief description of the problem on the macro aggregation level by the facilitator based on several interviews and literature study in order to select participants

P1-3 Whole problem: a first global selection by the initiator(s) and/or the facilitator of the aggregation level and the scope from which the problem will be viewed, based on the mental idea of the problem, of the initiator and/or facilitator

P1-4 Domain knowledge: the selection and invitation of the participants and their reference groups[25] according to the selected aggregation level and scope, based on the mental idea of the problem, by the initiator and/or facilitator

P1-5 Whole problem: a brief description of the problem by the facilitator according to the macro aggregation level of the problem. And an

P1-6 General preparation: an explanation of the method and tools by the facilitator

P1-7 Domain knowledge: a description of the professional background and expertise of the participants

P1-8 Domain knowledge: a description by the participants of the domain knowledge and data of the problem

P1-10 Whole problem: a broad description by the facilitator of the whole problem based on all the material, with a special attention to the

P1-11 Whole problem: responses of the participants to the description of the whole problem

P1-13 Whole problem: combination all information, with a special attention for the contradictions in the material

The product of the problem handling process of the Aids problem of the problem handling process P1-2 is a global description of the whole problem.

The Aids problem is an actual complex interdisciplinary societal problem, which has great implications on society, although the impact of the consequences of the Aids disease varies in different countries and in different parts of society.

Africa was one of the first countries in which people became infected with HIV[26]. Large parts of Africa, already damaged by poverty, war and tropical diseases, suffer great human losses as a result of the disease. In some African countries a significant part of the population is infected with the virus, is ill or has already died. Many of these men and women are between the age of 15 - 45, the age that is responsible for keeping the society going, for giving birth, educating the youngsters, farming the land and taking care of the economy. The loss of large parts of this active population has a great impact on African society.

The spreading of the virus in the rich Western European and Anglo-American countries started among homosexual men[27], intravenous drug users and hemophilia patients[28] (Chin, 1990)[29].

In some parts of the world, such as Asia, for instance in India and Thailand[30], there were few Aids victims reported until 1990-1991. Now there are signals that these countries will catch up rather rapidly[31]. The prognosis for the spreading of the Aids disease in Thailand will be millions of people in the next decade (Daily Bangkok, August 1990).

The virus causes the decline of the immune deficiency system which after some period of severe illness will lead to death. Because of the long incubation time in which the host of the virus is often not aware of being infected, there are many chances that the virus can be transmitted to other people. The disease can easily spread through sexual contact, because many sexual contacts are without protection. Due to the way the disease is spread by sexual activities and needle sharing among intravenous drug users, the victims are those people, upon whom because of their age, the economics of a society depends. In the western part of the world the first victims were mainly men, many of them unmarried. Gradually, the virus spread among intravenous drug users, prostitutes and women. Many of the infected women have, or will get children. The newborn children can be infected by their infected mothers, while their other children will be orphaned at a very early age. Aids is one of the major tragedies of the last decades of the 20th century and will probably remain a tragedy throughout the first decades of the 21st century.

Starting as a new and unexpected problem, the Aids problem confronts mankind with the notion that all the scientific and technological knowledge that is available is insufficient to provide an adequate answer to the problem.

In the western world most people are no longer used to being confronted with a deadly, transmittable disease. This is one of the reasons why people are so shocked by the Aids problem. In the Western world the idea had taken root that all diseases could be cured[32]. The general opinion is that terrible, deadly diseases only occur in tropical countries or is a problem of elderly people. We soon forgot that in the first half of the twentieth century there were many fatal diseases, such as tuberculoses, Spanish influenza[33], syphilis, polio while even measles caused the death of many people. Now many of these diseases can be cured with antibiotics[34].

In the example of the aids problem we assume that the scope in which the problem will be viewed is the spreading of Aids through the homosexual population in Amsterdam. The aggregation level at which the problem will be viewed is the macro level. The goal of the problem handling process could be to answer the question: "How is the virus spread among this population?"and "What is the effect on the homosexual way of life in this area?".

Within the disciplines of medicine, psychology, sociology, education, economy, law, politics, and theology, knowledge of the phenomena held by experts of different professions, to name a few:

the district nurse (Eggenkamp, 1989), the family doctor (Heitkamp, 1989; Wigersma, 1989; Meijer, 1990; Hoeksema, 1989), the cancer specialist (Beutler, 1987; Okusawa, 1988) the long specialist (Masur, 1978), the dermatologist, the venereologist, blood specialist, hemophilia specialist; it concerns issues that are studied in the domain of epidemiology, prenatal and post-natal care (Vedder, 1989), child care, and it is a concern to patients, physicians and researchers etc.;

welfare work (Heiden, 1989), buddy-project, intravenous drug users (Hartgers, 1989; Hoek, 1989; Coutinho, 1990; Ameijden, 1992), drug addicts, homosexuality, bi-sexuality, prostitution (Richters, Donovan, Gerofi & Watson, 1988; Kleinegris, 1990), child prostitution, death counselling, euthanasia[35], taking care of orphans, family life;

education in schools (Bezemer, 1989; Seydel, 1989; Wafelbakker, 1989), sex education to adults, special groups like drugs users, prostitutes and homosexuals; it concerns the domain of child care;

the insurance companies (Raadt, 1989), patients, employees (Ven, 1989), police force, financial support for medical care and education;

the labor market, the work powers (Roscam Abbing, 1989), the cost for patients, cost for hospital buildings, condoms production, marketing and distribution, financial support[36], police force;

planning of care of Aids patients by the (local) government (Scherphuis, 1988; Broekhuizen, 1989; VNG, 1989; NCOG, 1989; PRVN, 1990), the policy on AIDS and drugs (WVC, 1989; NCAB, 1990), developing organizational networks (Tillemans, 1988);

immigrate laws (USA[37]), laws for protection, labor protection laws (Boer & Lagaaij, 1989), guide-lines for prisons, guide-lines for criminal behavior, national and international law (Kastelein,1989), the protection of privacy of the patients (Bowles, 1989), fundamental rights (Frankena & de Graaf, 1990);

education on television (Condition Critical, ), documentaries with special topics on Aids (The Plague, ), covering special moments, such as Aids memorial, radio discussion with specialists, journals, books, ego documents (Jepson-Young, 1993).

families (Katwijk, 1989; Ossendrijver-Hogerhorst, 1989; Jepson-Young, 1993) including wives, lovers (Cohen, 1990), children, sisters, brothers, mothers[39], fathers, colleagues, friends (Ossendrijver-Hogerhorst, 1989), neighbors, doctors, care-takers, school teachers, lawyers, judges, policy makers, researchers, social workers, buddies, prostitutes and pimps, teachers, undertakers etc.

hospital buildings, hospital policy (Dutree, 1989), research institutes, universities (Dutch Program Committee for Aids Research, 1992).

The people who have studied the relevant disciplines, or who work in the fields mentioned above, have the knowledge of different parts of the problem. Each person has its own special knowledge and view of the problem. In order to get a single clear picture, all these people should exchange information with each other about the problem. However, to invite them all would make the problem handling team too large. In order to keep the problem handling process manageable there has to be a selection of experts. At this early stage of the problem handling process, the experts have to be selected before they themselves have had the chance to discuss thoroughly which domains are involved and what knowledge is needed to handle the problem. It may be necessary therefore to alter the composition of the team later on. The experts will be informed of this possibility and made aware that in case of expertise overlap, they may be asked to give up their seat in favour of another expert. However, as far as possible, a careful selection of the experts should avoid this unpleasant situation. The selection of experts should be made in view of the aggregation level chosen, the scope and goal of the problem handling process. Selection should be based on the main domains that are involved in the problem.

Another point to consider is whether all (or almost all) persons should be educated to the same level. When all the participants have, for instance, an academic degree (which will not always be possible) there is a kind of communication possible that can be indicated as a 'general academic way of discussing', even when the participants are from different disciplines. However, in society there are differences in social position. It may be necessary to include a wider variety of viewpoints. For instance, in case of the Aids problem, the educational aspects can be perceived differently by people from different backgrounds and belonging to different social levels.

These aspects have to be considered in view of the aim of the problem handling process[40].

The team could consist of[41] men and women with expertise in the domain of a family doctor, a medical specialist, a medical aids researcher, a buddy, a social worker, a homosexual aids patient, a HIV positive person, a nurse, a hospital director, an economist, a lawyer, a sociologist, a psychologist, an educationalist, a member of a homosexual pressure group.

In this particular example of a complex interdisciplinary societal problem it should be possible to combine several aspects in one person, if the team threatens to become too large, for example a HIV positive medical specialist or an educationalist with Aids.

The product of the problem handling process P1-8 contains domain knowledge. In this example illustrations are restricted to the descriptions of two domains out of the whole range of domains that are involved with of the complex interdisciplinary societal problem Aids. We have selected the domain of Aids as a medical problem and Aids as a social problem. We begin with Aids as a medical problem.

A description of medical aspects of the Aids problem could contain the following aspects: what is Aids, how is the disease transmitted, what kind of prevention can be implemented, what is the effect of prevention?

Aids refers to a combination of diseases, caused by the Human Immune deficiency Virus (HIV)[42], a virus that attacks, and in the end destroys, the immune system. The diseases resulting from the decline of the immune system are called Aids. Aids stands for Acquired Immune Deficiency Syndrome[43], a not innate deficiency of the immune system.

The task of the immune system is to defend the body against the attacks of viruses and microbes, and in this way prevent the body from acquiring diseases and getting ill. When the immune system declines, which can be temporarily at moments of great fatigue or in case of a disease, or permanently in the case of being infected with HIV[44], the body is unable to fight the viruses or microbes optimally. The more the immune system declines the less it can protect the body from diseases[45]. The attack by viruses and microbes can come from outside, as for instance the influenza virus, the measles virus or where there is a wound, or the attack can come from inside, in which case the virus and microbes are already within the body[46]. A body contains all kind of viruses and microbes which in normal circumstances, with a normal immune system, cause little harm. However when the immune system declines these viruses and microbes can cause severe diseases. Some of these diseases are called opportunistic infections. Examples of opportunistic infection are cold sore, pneumocystis carinii pneumonia, a special kind of pneumonia and herpes (Shilts, 1987).

When the decline of the immune system is caused by HIV and the person has (a combination of) certain special diseases that are characteristic of Aids, then the person has Aids (Stienstra, 1990, pp. 437-438). There has been a long discussion about a clear diagnosis of Aids. The following definition is accepted:

Three composed criteria are given for the diagnosis of Aids by CDC[47] ( 1987) and GHI[48] (1987) (Steering Committee on Future Health Scenarios,1992, p. 17).

I Pneumocystis carinii pneumonia in combination with a negative HIV-test result if no other causes .....are present.

II ...a doubtful laboratory HIV-test result (or absence of a test result) and no other causes for the immune deficiency ....are present. In combination with one of the indicators of Aids:

.........,specific types of candidiasis, ......., ......, ......, cryptococcosis, cryptosporidiosis with diarrhoea persisting >1 month, specified consequences of herpes simplex virus infection, kaposi's sarcoma affecting a patient < 60 years of age, lymphoma of the brain with a patient < 60 years of age, pneumocystis carinii pneumonia, toxoplasmosis of the brain affecting a patient > 1 month of age.

III ...a laboratory evidence of HIV-1 infection or a combination of following diseases..

...' HIV encefalopathy (also called Aids dementia of subacute encephalitis due to HIV)[49]....extrapulmonary tuberculosis by M. tuberculosis, ......, ....., HIV wastingsyndroom ('slim disease') and the indications under II."(Steering Committee on Future Health Scenarios,1992, p.17 , 18)[50].

"Strictly speaking, Aids,..., is not the name of an illness at all. It is the name of a medical condition, whose consequences are a spectrum of illness....., the very definition of Aids requires the presence of other illness, so-called opportunistic infections and malignancies. But though not in that sense a single disease, Aids lends itself to being regarded as one - in part because, unlike cancer and like syphilis, it is thought to have a single cause."

Like many other diseases caused by virus, the virus is transmittable. Fortunately the transmission is not as easy as with measles, or with influenza, where a huge population can be infected within only a few months. A person infected by HIV is called HIV-positive. The moment of infection is called sero-conversion[51].

Parts of the domain description of the medical specialist, combined with the description of the medical aids researcher and the family doctor may contain the following aspects:

1 The most frequent way of sero-conversion is by sexual contact, by means of blood-blood, blood-sperm, sperm-mucous membrane and blood-mucous membrane contact, sexual fluids and vaginal mucous membrane[52].

The spreading of the infection in the western world started among homosexual men. It spread rapidly in circles of homosexual men with frequent change of sexual partners (Shilts, 1987; Dondero & Wilson, 1992; Gill & Tillett, 1992). Via bi-sexual men the virus is spread to the heterosexual population[53]. In the developing countries the spreading of the virus is mostly via heterosexual contacts, often via prostitutes and promiscuous men.

2 Sero-conversion by blood-blood contact can happen by using already used hypodermic needles. When there is infected blood on the needle the next user of that needle will be infected too. In this way the disease can be spread rapidly among the population of intravenous drugs users (IDU) and can be spread via prostitution to other populations (Hartgers, 1991).

"In the USA, there seems to be an increase in the risk of HIV infection due to increasing cocaine use. Cocaine injection has been shown to be related to risky injecting behavior and to HIV seropositivity. Smoking or inhaling cocaine appears to be related to risky sexual behavior and to HIV infection." (Dutch Program Committee for Aids Research, 1992, p. 39).

3 An other way of sero-conversion by blood-blood contact is due to medical activities.

- In the cases where one uses instruments that are not (properly) sterilized so that there is a chance that infected blood on the instruments comes into contact with the blood of an other person. This especially concerns needles and instruments in surgery.

- By getting a blood transfusion with infected blood. In most western countries the blood is treated before it is distributed. There have been cases in which unchecked blood has been used, even when the possibility to check the blood was already available, for example in France in early 1986[54] and in some hospitals in Germany even still in 1993, in Egypt, in tropical countries, and in Eastern-Europe for example in Rumania.

Sero-conversion 1 till 3 are called horizontal conversion. Sero-conversion via mother and new born child is called vertical conversion (Newell, Dunn, Peckham, Ades, Pardi & Semprini, 1992).

The chances of getting infected through contact with infected blood or sperm, via blood or mucous membrane are estimated at 100%[55]. The chances that a mother infects her new born child are different in several areas. In Western countries there is a chance of 30% and in African countries this chance is 50%. What exactly plays a role in causing this difference is not certain[56].

The infection can be transmitted from the moment of sero-conversion until death[57].

The period from the moment of sero-conversion and actually becoming ill, which is called the incubation period, can be a period which can vary from three years to more than twelve years, as far as it is known now[58]. This incubation time varies in different populations, and varies also from person to person within the population[59].

There is little known, at this moment (May 1994), about the reasons for differences in incubation time, nor about the reasons for the difference in length of the actual disease. The mean/average time between sero-conversion and the outburst of the disease is at this moment nine years in the Netherlands. From becoming ill, what is called getting Aids, till death, takes approximately two and a half years.

The following description could be derived from the family doctor, combined with knowledge from the medical specialist, the medical aids researcher, the social worker, the HIV positive person and the Aids patient.

-to prevent the body with a vaccination in order to develop enough anti-bodies which can fight the disease in case of infection

-curing the disease[61]. Curing can be done by rest or in a combination of rest and medical care like a operation and/or medicines.

HIV conversion can be prevented if it is possible to find a fluid that can be given to a person to protect against the virus, by forming antibodies that fight the virus. In this way the body is protected in case of infection and prevented from becoming ill. However a good vaccinehas sofar not been developed (Meijer, Molema & Jansen, 1990a). One of the reasons that makes it very difficult to create a good functioning vaccine is due to a replica of the DNA-RNA string. By copying itself the virus is 'careless'. At the end of the DNA string there are some deviations in the copy compared to the original. These 'mistakes' make it very difficult to connect some kind of vaccine to the virus. This also ensures that each person is infected by a slightly different virus[62]. This makes is hard to develop a vaccine against the virus (Meijer, Molema & Jansen, 1990a, pp. 440-443; Back, Thiriart, Delers, Ramautarsing, Bruck & Goudsmit, 1990). At this moment it is not possible to prevent sero-conversions of HIV by vaccination (Boucher, Krone, Goudsmit, Meloen, Naylor, Goldstein, Sun & Sarin, 1990; Goudsmit, Back, Thiriart & Bruck,1990).

However, having a vaccins developed at this moment would only solve part of the Aids problem, because a vaccination only helps those people who are not yet infected. It has also proved to be very difficult to get a high vaccination rate in all countries especially in the developing countries where money and (sometimes) a good functioning infrastructure are missing.

In the case of sero-conversion the decline of the immune system is irreversible. Once infected, the decline of the immune system cannot be stopped or reversed. Being infected ends with death within three to fifteen years. Until now it is impossible to prevent the immune system from declining after infection, although medicines can to a certain extent lighten the diseases or slow down the decline, for instance by giving AZT[63] or DDI or both[64]. Effect of AZT/zidovudine is measured by the amount of CD4+ T-cell numbers[65] (Gruters, Terpstra, Lange, Roos, Harkema, Mulder, Wolf, Schellekens & Miedema, 1991; Dutch Program Committee for Aids Research,1992, p. 24).

"T-cell responsiveness and the number of CD4 T-cells improved six months after the start of zidovudine treatment, but only for a short period."

The only way to prevent the body from getting infected at this moment is to prevent infection.

The main preventions of sero-conversion can be distinguished by different ways of distribution

Prevention of sero-conversion by sexual contacts can be achieved by abstaining from all sexual contacts or having a 'controlled[66]' monogamous relationship.

A way to reduce the danger is having sexual contacts with only a few persons or using adequate protection to prevent sero-conversion. This can be done by using condoms. There are condoms for men and recently also for women.

2 Prevention of infection through infected needles of, for instance, intravenous drug users.

Prevention of infection through infected needles can be done by using sterilized needles for each injection.

Prevention of infection of medical instruments can be done by using sterilized instruments for each medical activity.

Preventing sero-conversion via blood transfusion can be ensured by giving blood a special treatment before using it.

Until now the medical world has been unable to prevent that a mother infecting her foetus or her new born child. The only way to protect against this kind of infection is to prevent the pregnancy of HIV infected women.

6 Other ways of preventing the spread of the virus are more related to the political or juridical level.

Juridical measurements, such as for instance in the USA where no immigration of HIV-positive people is allowed. The USA closing her borders was the reason that the 7the International Conference on AIDS was held in Amsterdam, the Netherlands instead of in Boston. Sweden and Cuba did put some HIV-positive people in isolation camps.

The following description could come from the family doctor, combined with knowledge from the medical specialist, the medical aids researcher, the social worker, the HIV positive person, the Aids patient, the psychologist the sociologist and the educationalist.

Having no sexual contacts or only one monogamous contact would be a solution. However, this is not a realistic means of prevention.

In addition to approaches to prevention based on medical activities, much energy has been put into educational campaigns for preventing sero-conversion through sexual contacts by using condoms. However the campaigns for using condoms to prevent sero-conversion are mainly focussed on, and appeal to cognitive arguments and to cognitive knowledge. At the cognitive level the campaign has been successful, although less successful at the behavioral level, so that the combined result was less positive than had been desirable[67].

"A sharp decrease was observed in all registered STDs among homosexual men. The strong similarity between the incidence of these infections and the incidence of HIV-1 infection in the cohort showed that the process of sexual behavior changes among homosexual men in Amsterdam was representative for the homosexual community in Amsterdam. It appeared that the incidence of syphilis tends to reflect behavioral changes sooner than the incidence of hepatitis B. The rise of incidence of HIV-1 infection observed in 1990 (Table I) prompted us to investigate whether this was caused by a rebound in sexual behavior. It appeared that both the frequency of anogenital receptive intercourse and the number of partners had increased among recently infected men as compared with men who had remained seronegative."

2 Preventing drug users from using the needle of another person is not so difficult. In the Netherlands there is a rather successful action of a needle exchange program (Cohen, 1990; Hartgers, van den Hoek, Krijnen, Coutinho, 1992).

"In Amsterdam a special approach towards drug users, called 'harm reduction' has been developed. The goal of harm reduction is "to create a situation that greatly reduces the risk of addicts harming themselves or their environment". As part of this approach "low-treshhold" methadone programs and a large-scale needle/syringe exchange program have been implemented." (Dutch Program Committee for Aids Research, 1992, p. 39)

At several places in Amsterdam drug addicts can exchange used needles for a new sterilized one. This slows the spread among intervenes drug users, but does not stop it. This idea has been followed as an experiment in France and Germany in some parts of USA (Cohen, 1990).

3 In most of the western countries, sero-conversion as a result of medical activities, like using needles and blood, is prevented by the use of new sterilized instruments for each medical activity. In other countries, for instance some of the Eastern-European countries and developing countries, this procedure of using sterilized instruments is too expensive.

4 Sero-conversion caused by blood transfusions can be prevented by specially treating the blood before using it, it seems to be quite under control in the western countries[68]. In the developing countries there is often no money for this treatment, however, and there are reasonable chances that blood in those countries is infected.

5 Sero-conversion of foetus or new born child from the mother happens during pregnancy or during the partus or shortly after the partus by breast feeding[69]. The only prevention for this vertical sero-conversion is to discourage HIV-positive women from getting pregnant or after pregnancy encourage them to get an abortion[70].

6 Prevention of the spread of the virus by putting infected people in isolation camps is not only cruel, but also based on the incorrect idea that it is possible to select and isolate all HIV infected people and in this way prevent the population from having contact with infected people. Many infected people do not know that they are sero positive, nor are they known to the government. Nor can the government prevent their citizens meeting people from abroad who may be HIV infected. Not allowing HIV infected people to enter the country is probably a measure taken not only to prevent the spread of the infection, but also to reduce health care costs.

Discussing the so-called medical aspects of the disease makes it clear that these are not purely medical aspects. The medical aspects cannot be isolated from other aspects of the problem, but are for instance embedded in an economical, sociological, psychological, cultural, ethical and theological context. Even a rather simple medical aspect of vaccination is embedded in economical aspects: "Is it worthwhile to spend money to develop a vaccines?", "Can people afford to buy it?". There are also some psychological aspects, and even theological aspects: "Are people willing to use it?"[71].

- domain knowledge: an explanation of domain knowledge and data by the participants

G1-2 General preparation: the explanation of the method, and tools including the function of the seven layer model

G1-3 Whole problem: a short discussion of the description of the problem by the facilitator and a discussion about the selected aggregation level,

the demarcated scope and the way the aggregation level and the scope should be approached

G1-4 Domain knowledge: an explanation of the domain descriptions of the problem by the participants

G1-8 Whole problem: a first attempt of a description of the whole problem at the macro aggregation level on layer I including the description of which domains, what concepts and which phenomena are involved

G1-9 Whole problem: a discussion of "What have we forgotten?". The participants are given new assignments for preparation and are asked to formulate new questions

At the end of session one there is a global description of the whole Aids problem at the macro aggregation level based on the domain description of the problem by the participants. A start is made with a description of the domain knowledge of the whole problem in layer I. The first attempts at a description of the whole problem could, as well as the aspects we have already described, include the following, derived from the descriptions of the social worker, the HIV positive person, the Aids patient, the psychologist, the undertaker.

In this first group session the participants meet each other for the first time. Not many people are used to discussing problems in a multi disciplinary team and to exchanging information on equal level with persons from different hierarchical societal levels. In the beginning this may be perceived as rather strange and it may invoke uneasy feelings. The facilitator should at all times take care of an equal contribution of the participants[72].

From the description of the social worker, the psychologist and the sociologist, information could be reported about HIV infected women, pregnancies and the effect of the infection on the new born and already born children. These people could also report on cultural changes, for instance concerning funerals.

The death of young homosexual men, who die of Aids has caused a gradual deveopment of a new funeral and burial culture. The funerals of homosexual men are often prepared in advance by themselves, and organized by their closest friends. Some of the funerals are real happenings, more a kind of farewell feast. It is regarded as the last means of the person to express himself, which is quite different from the usual funerals in the Netherlands.

Another change in culture within the homosexual community may be the safe sex parties, the HIV positive cafe's and the action groups as, for instance, Fight for Life and Act Up.

From the descriptions of the social worker, the psychologist, the HIV positive person and the Aids patient, information could be reported about the way in which families and friends react to having a family member or friend who is HIV infected or has Aids.

From the educationalist's description, information might be reported on educational programs for intravenous drug users, prostitutes, and brothel keepers.

From the economists account together with that of the hospital director, data might be assembled on hospital costs, costs of medicines, the benefit of producing pharmaca for Aids patients, the costs of hospitalization, the medical costs of a single patient, the nursing costs, the costs for training new employees as well as the costs in loss of life expressed in the loss of an educated adult, a member of society.

Together with the psychologist and the sociologist the economistcould try to calculate these costs of loss of life. This may be expressed, for example, in the cost of education provided, absenteism from work, the costs of loss to family and friends, and the costs regarding the orphans.

From the knowledge of a lawyer, data could be derived about the rights of patients, the rights of an employee with HIV or Aids, the rights of an insurance company versus the rights of a HIV infected person or a person with Aids. The issue of testing (anonymously) of all people for HIV prevalence, the pension payments of HIV positives. The fights of heamophilia patients infected with infected blood, after it was already known that blood could be infected and tests were developed.

The buddy, the HIV positive person and the Aids patient can all provide information about social abuse, social isolation and dismissals from work and rejection from the labour market.

- Whole problem: To have a list of the concepts and the phenomena involved in the problem

- Domain knowledge: To have a group of participants composed according to the knowledge that is required

- Whole problem: To have a description of the theoretical ideas about the connections between the concepts with each other and the phenomena

- Domain knowledge: To have an inventory of what the team knows about the problem and what is still unknown and unclear

- Whole problem: To have a semantic model of the major connections between the concepts and the phenomena at the macro level of the whole problem

Data gathering related to the concepts and the phenomena according to the firmer mental idea of the problem and hypotheses being formed about the problem.

P2-1 Domain knowledge: New data about the domains of the problem by the participants

P2-2 Whole problem: An inventory by the participants of the concepts and phenomena that are involved in the whole problem

P2-3 Whole problem: A combination and categorization of concepts and phenomena of the whole problem by the facilitator

P2-5 Domain knowledge: The relation between the concepts and the phenomena to their own expertise by the participants

P2-6 Domain knowledge: A combination of these lists of concepts and phenomena and expertise of the participants

On the basis of the described domain knowledge and the combined knowledge, the participants could, either separately with their reference group or together with other participants, extend the description of the whole problem by reference to the field they have some knowledge about.

Because the group is selected with a view to the ultimate goal of the problem handling process, concerning the whole problem, it is likely that some of the knowledge is missing. For instance, knowledge of the growth of the disease in other parts of the world such as Africa, Asia and South-America or knowledge about drug users in regard to medical, economic and social aspects. The participants may now be asked to assemble data concerning the whole scope of the problem, in accordance with their own domain, in order to make the overview of the whole problem more complete. This could include some information about the economic and social consequences of the disease in other countries than the western countries, or information about the reaction of the World Health Organization or the World Bank to the Aids problem.

- whole problem: a discussion of the list of concepts and phenomena of the whole problem

- domain knowledge: a discussion of the expertise of the participants in connection with double or missing expertise concerning the phenomena and concepts

- whole problem: a semantic model of the major connections between the concepts of the whole problem

G2-2 Whole problem: a discussion of the list of concepts and phenomena of the whole problem

G2-3 Domain knowledge: a discussion of the duplicated and the lacking expertise, and the team consequences

G2-5 Whole problem: the concepts and the phenomena are included in the description of the whole problem

G2-6 Whole problem: an update of the list of the concepts and the phenomena of the whole problem

G2-7 Whole problem: the definition of the concepts and the phenomena of the whole problem; constructing layer II

G2-8 Whole problem: a first attempt describing the theoretical ideas of the whole problem; constructing layer III

G2-9 Whole problem: hypotheses about connections between concepts and phenomena of the whole problem; constructing layer V

G2-10 Whole problem: a semantic model based on the list of concepts and phenomena of the whole problem

G2-11 Whole problem: an update of the descriptions of the whole problem in the layers I, II, III and V

G2-12 Whole problem: a discussion of "What have we forgotten?". The participants are given new assignments for preparation and are asked to formulate new questions.

The list of concepts and phenomena compared with the knowledge of the participants

- what can be done about taking care of the orphaned children whose mothers have died of Aids.

- how should we take care of children of mothers with Aids and who should take care of these children (Congress Women and Aids, 1990).

- what should be the policy of the government with regard to abortion in the case of HIV positive mothers.

- what are the legal rights of an infected hemophilia patients (Nederlandse Vereniging van Hemofilie-Patienten, 1986).

- what are the reasons why some people have a shorter incubation time than others

- uncertainty about the effect is of education (Coutinho, Van Griensven & Moss, 1989; De Vroome, 1994).

- uncertainty about how the virus originated. There are different theories about this. Is this relevant for analyzing the problem (Sabatier, 1988)?

- how many people are dismissed or not hired because of HIV or Aids (Van de Elsen, 1994).

- why there is a difference in latency period, the specialist can than point to research of Dutch Program Committee for Aids Research (1992, p. 16)

- what is the effect of education in different groups in the Netherlands, in Africa, in Thailand.

- what is the economic effect of absebteeism, orphanizing, intensive care of Aids patients

The relevant questions are selected for further analysis. The less relevant questions are, temporarily, set aside.

On the basis of this discussion, a first semantic model of the concepts and the phenomena involved in the whole problem can be made.

- whole problem: a causal model of the whole problem and an update of the layers I, II, III, V

Layer II : whole problem: the second definition of the concepts and the description of the phenomena of the whole problem

Layer III : whole problem: the second description of theoretical ideas of the whole problem

P3-1 General preparation: the facilitator will eventually invite ne participants and introduce them to the working process

P3-2 Domain knowledge: an update of the domain description of the problem and the definitions of the concepts and the phenomena by the participants together with their reference group

P3-3 Domain knowledge: a first description of the theoretical ideas of the domain knowledge by the participants together with their reference group

P3-6 Whole problem: an indication of causes and effects of the phenomena according to the macro level of the whole problem

A more extended and more detailed description of the different aspects of the Aids problem in relation to their specific domain is now provided by the participants, including a definition of the domain concepts and a description of the phenomena. This is followed by a description of the theoretical ideas in the domain by the participants together with their reference group. For participants from the medical domain this may include hypotheses about the spreading of Aids in the western world in the different populations. Based on interviews or the analysis of data from epidemiologists, different risk groups could be defined and hypotheses formulated on how the virus has spread, and will spread, among these groups as well as on how the different groups are connected. One hypothesis could be that the virus has spread from homosexual male group via bi-sexual men into the heterosexual population. Another hypothesis could be that the virus has spread from intravenous drug users via prostitutes into the heterosexual population. A hypothesis about the economic consequences of Aids in Africa could be that a whole area could be impoverished for decennia because the people who have to take care of the economy are either ill or dead. A hypothesis about Thailand might be that there is a connection between low economical areas (in the north) where young girls are sent to the south (Bangkok, Phuket) to work as prostitutes where they are infected by their clients. Another hypothesis could be that the way of life of homosexual men will change, partly because many creative persons in the (sub) culture died, for a part because they will in a way adjust to this new situation. In the medical domain there are already are some proven and tested theories about what the virus looks like and about the way the virus replicates itself. However, there are many issues still unclear. For instance, it could be that in the case of a shorter incubation period, the whole constitution of a person plays a role. There may be theories in the field of psychology about short term and long term goals. There are also some theories of Freud concerning the principle of lust. In the field of sociology theories on finding a (sexual) partner with a reference or peer group may be of use. The nurse could have ideas about family reactions towards family members with HIV or Aids etc.

- whole problem: a discussion of the description of the whole problem, the concepts, the phenomena and the theoretical ideas based on the knowledge gathered in the previous sessions.

- whole problem: a construction of the causal model of the whole problem based on the semantic model and the list of cause and effects.

G3-2 Domain knowledge: a discussion of the updated domain knowledge of the participants

G3-3 Whole problem: an update of all the descriptions in layers I, II, III, V of the whole problem.

G3-4 Whole problem: a discussion of the theoretical ideas of the whole problem and of the domain knowledge in the light of the question list

G3-5 Whole problem: a discussion about the list of causes and effects of the whole problem, made by the participants

G3-6 Whole problem: a discussion of whether a causal model of the problem is fruitful

G3-8 Whole problem: a formulation of some hypotheses about cause-effect connections between phenomena of the whole problem

G3-9 Whole problem: a causal description of the whole problem in small sub- groups

G3-10 Whole problem: a plenary discussion of the causal descriptions of the whole problem

G3-12 Whole problem: a plenary discussion about the causal model of the whole problem

G3-17 Whole problem: a discussion of "What have we forgotten?". The participants are given new assignments for preparation and are asked to formulate new questions.

Now a causal model based on the former semantic model can be given of the whole problem of the Aids problem. Based on the semantic model, relations can be indicated in the form of cause and effect, a '+' between a and b means that when a increases b increases too; a '-' between a and b means that if a increases, b decreases.

Layer II : domain knowledge: the definition of the domain concepts and phenomena

Layer IV : domain knowledge: a description of the domain knowledge islands Layer V : domain knowledge: the semantic domain model

P4-3 Domain knowledge: on the basis of new knowledge and data the participants are asked together with their reference group to implement the domain knowledge in layer I, II, III. The participants are assisted by the facilitator

P4-5 Domain knowledge: on the basis of descriptions of the domains reference group will make a first semantic model of the domain knowledge.

P4-7 Whole problem: a connection of the domain layers with the layers of the whole problem, done by the facilitator

The description of the knowledge islands may show that there is a reasonable amount of knowledge, although not enough, concerning the virus and the disease and the spreading of the virus in the homosexual community compared to the knowledge of other fields. Some areas in the medical field are well studied and some areas of the field of epidemiology are also studied well. These knowledge fields lie parallel with the research fields that are funded (see Dutch program Commitee for Aids research, 1992). In figure 13 the way the size of the cycles, and the way they are shaded indicate the existing knowledge and the knowledge that is required for analysis of the problem. The more shaded the island is, the more knowledge is available. The white areas in the model indicate the the blind spots. The white cycles, or almost white circles indicate the white spots, the issues we know nothing or little about.

G4-3 Whole problem: an update of the models of the whole problem in layers I, II, III, IV and V based on new ideas and insights of the whole problem derived from the domain knowledge.

G4-4 Whole problem: a discussion of the theoretical ideas about the whole problem

G4-5 Whole problem: a discussion and update of the causal model of the whole problem including the latest changes

G4-7 Whole problem: a discussion based on the question "What have we forgotten?". The participants are given new assignments for preparation and are asked to formulate new questions.

A product of the problem handing process G4-7 concerning the question "What have we forgotten?" could be:

At this moment in the discussion the participants might realize they had forgotten to analyze the influence of is the spread of the virus in the East-European countries or the effect of people with Aids in the developing countries on their health care organizations, or the effects of the people that died of Aids on the economy of their country.

- whole problem: a careful description of the theoretical ideas on which the connection between the concepts and the phenomena of the whole problem are based

- whole problem: a selection of the aggregation level and scope of the whole problem

P5-1 Domain knowledge: the participants are invited together with their reference group to criticize the domain knowledge in layer I to V and to reflect the domain theory

P5-3 Domain knowledge: the participants are invited together with their reference group to describe the theoretical domain ideas

P5-4 Domain knowledge: the participants are invited together with their reference group to make a list of the concepts and the phenomena, indicating causes and effects, of their domain

P5-5 Domain knowledge: the participants are invited together with their reference group to make a description of the domain knowledge indicating causes and effects

P5-6 Domain knowledge: the participants are invited together with their reference group to make a causal model of the domain knowledge, in which they are assisted by the facilitator

P5-7 Whole problem: a critical discussion of all the layers of the whole problem of the participants with their reference groups

P5-10 Whole problem: the facilitator combines all the critique and send the reviews to all the participants

In reflecting the domain theory, psychologists might consider the connection between a famous person dying of Aids, for instance the well-known movie star Rock Hudson in 1985 (The Plague) or the tennis player Arthur Ashe, as a HIV positive person, and the acceptance and awareness of the disease.

The actually description of the theoretical domain ideas, for instance by an economist, could lead to hypotheses about relations between individual loss of money and loss of money at the national level due to Aids. For instance, the financial costs of young men dying of Aids and hypotheses about financial models at the level of the world economy.

- domain knowledge: an explanation and discussion of all the models of the domain

- whole problem: a critical review of the models of the whole problem by external expert(s)

- whole problem: a selection of the aggregation level and scope of the whole problem

G5-5 Whole problem: an outsider is invited to play the role of the devil's advocate, to ask critical questions and give a critical review of the description of the whole problems in terms of the layers

G5-6 Whole problem: a discussion of this criticism, using brainstorming facilities

G5-7 Whole problem: an adaptation of new ideas based on the discussion of the layers I to VI of the whole problem

G5-8 Whole problem: a careful registration of the knowledge of the problem, what do we know (based on theory) and what are still hypotheses, what are assumptions etc. What is the (temporary) answer to a question and which questions remain

G5-9 Whole problem: a selection of the aggregation level and scope of the whole problem

G5-10 Whole problem: a discussion of "What have we forgotten?". The participants are given new assignments for preparation and are asked to formulate new questions

The critical review of the descriptions of the problem could lead to more or altered connections between the concepts and the phenomena, and could lead to many new questions. For instance, what should be the relation between education and the knowledge that during the incubation period there seems to be a period where the transmission of the virus is lower than at other perods in the incubation period (Dangerfield, 1990a, b).

A discussion about the selected aggregation level and the demarcated scope. The discussion could be about, for instance, in what way the homosexual population with many changing partners is a more or less 'closed' population and how much and in what way it differs from other populations. In this discussion the aggregation level and the scope must be carefully demarcated.

- Whole problem: an update of the models of the whole problem in layer I till VI

- domain knowledge: the filling of some of the white spots in the domain knowledge

P6-1 Domain knowledge: the participants are asked to evaluate their domain models together with their reference group

P6-2 Domain knowledge: the participants are asked to find data together with their reference group to fill in the white spots

P6-3 Whole problem: the participants are asked together with their reference group to find data that contradict with the description of the model of

With regarding to the problem again, especially the knowledge islands, it could be that there is too little knowledge about the medical aspects about women and Aids. It seems that the signals of HIV infection and the incubation period differ compared to those of men. There is also insufficient knowledge about the social and psychological aspects of children of mothers with Aids. The economic aspects of the illness and the death of mothers should be included too, as well as the health care costs for HIV infected babies. The domain experts could find out more about these issues in the literature, in journals and in reports.

- whole problem: starting to build the models again, now for the chosen aggregation level and scope of the whole problem

G6-2 Whole problem: an adjustment of the description of the whole problem accomodate the new ideas

G6-4 Whole problem: second evaluation of the models of the whole problem, while some of the participants play the role of the devil's advocate;

domain knowledge: a second evaluation of the domain models while some of the participants play the role of the devil's advocate

G6-7 Whole problem: a discussion of "What have we forgotten?". The participants are given new assignments for preparation and are asked to formulate new questions.

A product of the problem handing process P6-2 coul be the filling of some of the blind spots.

The group can be aware, at this moment, that no matter how complicated the problem already looks, there will probably exist many blind spots. It could be well-known knowledge that has just been overlooked, or it could be a white spot. Some of the blind spots that have been overlooked could be the brothels in Amsterdam where women (often mothers) from such countries as, for instance, the Philipines, work, and later go back infected to their country. A blind spot could be the risk of infection by way of incest and rape. The paradox is that as soon as the blind spots are found, they are no longer blind spots.

In the case of a very important problem, two teams can analyze the problem in parallel. After the first attempt of both groups to formulate the conceptual model, the facilitator will send the material of each group to the other group for preparation.

Repeating the sessions two to six, now for the selected aggregation level and the selected scope of the whole problem and of the domain knowledge

A product of the problem handing process seven could be a more detailed and extended description of the knowledge about the part of the Aids problem that is selected. A product of the problem handing process seven could be a system dynamic model of the selected aggregation level and the demarcated scope.

Depending on the desired situation, the focus might be the spreading of the disease through the population. In the epidemiology which is concerned with the way viruses spread through a selected population, distinctions are made between different groups of people, based on the idea that the speed of the distribution of infection and the amount of infected people differ significantly in different groups[74]. Some of the populations distinguished by Aids epidemiologists are, for instance, the male homosexual community of sexually active men with many different sexual partners; bi-sexual men with many different sexual partners; the intravenous drugs users; female and male prostitutes; hemophilia patients; the (young) promiscuous heterosexuals.

To construct a simulation model of the spread of the virus through the population, the participants can start with carefully defining the population and the demarcated scopes. The population selected is the sexually active (young) homosexual men with many various partners. The geographical scope is restricted to Amsterdam[75]. The time scope could be between 1984 and 1994. The main levels in this model could be: the number of people that are healthy, the number of people that are HIV positive, the number of people that have Aids, the number of people that died of Aids. In figure 14 a system dynamic model of the spread of HIV through the population is represented graphically. Unsafe sexual behavior can be defined on the basis of research on risk factors in the transmission of HIV and changes in sexual behavior.

Spreading of HIV through the population can be, among others, based on research data from the Dutch Institute of Social Sexuology Research (NISSO) (Dutch Program Committee for Aids Research, 1992, pp. 166-172; Sandfort, van Zessen, van Griensven, Staver & Tielman, 1991; Zessen, van & Sandfort, 1992). Their study is based on such questions as:

"1. In how far and along what lines are sexual contact networks build, and what is the contribution of those networks to the possible spread of HIV? In this perspective it is important to collect data not only on the characteristics of the respondents but also on those of their partners and on the process of partner selection.

2. What do interactional processes between sexual contact partners contribute to prevent behavior with regard to HIV? To answer this question, it is necessary, among other things, to investigate the negotiating processes between sexual contact partners.

3. How are sexual and preventive behavior actually produced by the individuals." (Dutch Program Committee for Aids Research, 1992, p. 166).

See for information and data about modeling Jager & Ruitenberg (1992), a workshop held at the National Institute of Public Health and Environmental Protection (RIVM); Dörner (1986); Heisterkamp, De Haan, Jager, Van Druten & Hendriks (1992); McCormick (1989); Druten, Jager, Heisterkamp, Poos, Coutinho, Dijkgraaf, Reintjes & Ruitenberg ( 1990); Druten & Jager (1991), and the Dutch Program Committee for Aids Research, 1992. p. 16, p. 19 refer to the studies of Wolf, Lange, Houweling, Mulder, Beemster, Schellekens, Coutinho, van der Noordaa & Goudsmit (1989) and Griensven, de Vroome, de Wolf, Goudsmit, Roos & Coutinho (1990).

"We estimated the prevalence of HIV-1 seropositivity among homosexual men in Amsterdam, by comparing the ratio of HIV-1 prevalence to AIDS incidence in the cohort with back-projections from the yearly number of new AIDS cases in Amsterdam. The number of HIV-1 infected homosexual men in Amsterdam was estimated to be between 1800 and 3500 ......"

"Data ..........came from sera obtained in a previous cohort study (1980-1982) of men with homosexual contacts, in whom the efficacy of a hepatitis B vaccine had been tested. Five of 685 participants (0.7%) had shown antibodies against HIV-1 on entry into the study, when their serum samples from the period November 1980 to December 1981 were tested. In 1981, three new cases of infection with HIV-1 appeared and during 1982, 12 seroconversions were detected. From these data it may be deduced that HIV-1 was introduced into the Amsterdam risk group of men with homosexual contacts at the end of the 1970s."

"...(the) new cohort of 741 homosexual men between the ages of 18 and 65 years who had had at least two different sexual partners during the preceding six months. In 233 (31,4%) participants, antibodies against HIV-1 could be demonstrated, which indicated an important increase in the number of infections in 1983 and 1984. The further spread of HIV-1 among study participants can be seen from the number of seroconversions that took lace after October 1984. Over a period of seven years, 109 new infections (seroconversions) were registered among 691 previously seronegative subjects. The annual and cumulative incidences of HIV-1 infection are shown in Table 1. A marked decline was observed in the annual incidence of new HIV-1 infections, from 8.9% in 1985 to 1.0% in 1889. It should be noted that the annual incidence of 1.0% in 1989 was followed by a rise to 2.9% in 1990 indicating a trend toward unsafe sexual behaviour. This trend was not limited to participants in the cohort study, however, as reflected by an increase in the number of new cases of gonorrhea and syphilis reported among homosexual men and bisexual men who visited Amsterdam's venereal diseases clinics."

In the Dutch Program Committee for Aids Research, 1992 is stated on p. 11 that mostly seroconverted after seven years to Aids.

".......the first case (of Aids) among IDU (intravenous drug users) was not diagnosed until 1985, by which time 50 cases had already been diagnosed among homosexual men. HIV was probably introduced among IDU by IDU who had homosexual contacts, although spread by IDU infected in other countries cannot be ruled out."

See for more research on Aids data and modeling also Haastrecht, van den Hoek, Mientjes & Coutinho (1992), Reinking, van den Boom, Jager & Postma (1989), Reinking, van den Boom, Postma & Jager (1992a) and Reinking, Postma, Alberts, van den Boom & Jager (1992b). See also the discussion on data concerning this kind of problems[76].

figure 15 Equations belonging to the System Dynamic model of a part of the Aids problem

On this point it could become clear that the model is not correct and that some important aspects are missing. The model as such is represented as a closed system, whereas in reality the population is not a closed population as is often assumed in the literature on epidemiological models of influenza (Metz, 1990). However, almost every group is an open population. For instance, in the population described above there was intensive interaction with the homosexual male groups in New York (Fire Island, The Plague, Shilts, 1987), Princetown (Cape Cod), San Francisco (Shilts, 1987). Of course the population is also open to the other side via bi-sexual men.

- whole problem: fill in some white spots, find some blind spots of the whole problem

- domain knowledge: fill in some white spots, find some blind spots of the domain knowledge

P8-1 Whole problem: the participants are asked to find data that conflicts with the models of the whole problem. This can be stimulated by making a new list with questions and doubts by the reference group

P8-2 Domain knowledge: the participants are asked to find data that is in conflict with the domain models

Difficult points in the discussion, concerning the global view of the whole problem as well as the demarcated scope of the problem, include the question of how detailed the data should be. Will it be enough just to indicate levels or will more specific data be needed. It will often happen that, as stated earlier, knowledge and data are not available, or at least not easy to get, or the data are in contradiction with each other. Sometimes it may be necessary to give a small assignment to someone who, within a short period of time, will search for some of the missing data. Another option is to conclude in the final report the fact that some of the necessary knowledge and data are still missing.

- whole problem: a discussion of the white spots and a discussion of the blind spots of the whole problem

- domain knowledge: a discussion of the white spots and a discussion of the blind spots of the domain knowledge

- whole problem: a connection of the selected aggregation level and scope of the problem to the whole problem

G8-2 Whole problem: fill in some white spots and discuss some blind spots according to the selected aggregation level and scope of the whole problem

G8-4 Whole problem: a discussion based on "What have we forgotten?". The participants are given new assignments for preparation and are asked to formulate new questions.

Blind spots in the demarcated scope could be the tourist population that visits Amsterdam and the connection between the homosexual population and intravenous drug users. Another blind spot could be the economic, cultural and emotional value of the early death of many young, promising artists who inspired others and created a new way of life and a new kind of culture.

With the knowledge of session seven and eight the group should go back to the whole problem at the macro aggregation level and alter these models according to the insights gained.

G9-1 Whole problem: broaden the level from the macro aggregation level and scope back again towards the whole problem

What is known now, at the chosen aggregation level and the demarcated scope, might have repercussions on what was said earlier about the whole problem. The new knowledge might make it necessary to change some of the presummed connections or aspects at the level of the whole problem.

Knowing that Aids changes the funeral and burial culture in the homosexual population the question should be examined of how this also changes the funeral and burial culture in the heterosexual population. Knowing what is known about the economic aspects of loss of homosexual life and loss of artistic life, some of the aspects of the economic models based on this knowledge have to be reviewed again.

P10-1 Domain knowledge: the participants are asked to discuss the models and the description of the domain problem together with the reference group in order to find blind spots

P10-2 Domain knowledge: the participants are asked to give a complete as possible domain description of the problem including all the domain models

For this discussion an outsider, who is also a domain expert[77], is invited to play the devil's advocate

G10-2 Domain knowledge: a discussion of the outcome of the domain discussion in finding blind spots

G10-3 Whole problem: a brainstorm in order to find new blind spots in the model of the whole problem

G10-4 Whole problem: a discussion of the brainstorming outcome concerning the blind spots of the whole problem

G10-5 Whole problem: an adaptation of the models of the whole problem according to the new insights

G10-6 Whole problem: a discussion about the consequences of the changes on the macro level and broad scope of the problem for the selected aggregation level and scope of the whole problem

G10-7 Whole problem: translation of this discussion to the models of the whole problem according to the aggregation level and the selected scope

G10-8 Whole problem: a discussion as to whether or not, and to what extent, the whole problem is by now satisfactorily analyzed and described

G10-10 Whole problem: a definition of the problem according to the selected aggregation level and scope of the whole problem

G10-11 Whole problem: a discussion of whether, with this description, the problem is defined

G10-12 Whole problem: a description of the whole problem according to the macro level and broad scope

The central issue in the discussion is that the problem of Aids is more complicated than it appeared at first sight. More issues are involved, more people and more areas. Another central issue is that there is little known about non-medical aspects of the problem. And as prevention fails, as it does in many cases, it will be very hard to find ways that can slow down the spread of the infection. Something that could be done is increasing the economic independency of women all over the world, especially in countries such as Africa and Asia (Thailand), in that way that women do not have to prostitute themselves in order to earn a living. A way of decreasing the spread of the infection would be to increase the standard of living of black women in the large cities in the USA, and to make it possible for them to combine raising children with earning money for their family in a decent way. This will make it possible for them and their children to get away from poverty, and the drug use which seem to have a direct relation with prostitution and HIV infection. Of course these issues are only suggestions for intervention and should be carefully discussed based on a thorough analysis of the problem in the second sub-cycle of problem handling.

It should become clear that until now the Aids problem has been defined too restrictedly to only some of the domains of the problem, such as the medical field and some aspects of social science, the educational parts and some areas of sociology, and the homosexual life. It may be clear now that the funds for research on Aids should be shuffeled, or extended, to other domains, such as the economic domain, the legal domain, the social sciences, in order to be able to explore the other effects and relations of the Aids problem. Especially the analysis of Aids as an interdisciplinary problem should be funded, since in this research the underlying coherence of the phenomena involved in the Aids problem can be studied.

- making a report including the method followed, the names of participants, their professions and their expertise, a global description of the whole problem and a description of the conceptual model of the problem according to the selected aggregation level and the scope, and a description of the domain knowledge. This report should contain the name of the problem owner and finish with the description of the remaining questions.

- an evaluation of the report by the participants together with their reference group

- an explanation by the facilitator of what is to be done with the results and how the further handling process will proceed

Finally, the final report[78] will be sent to all the participants and the members of their reference group. The participants and the refence group will be thanked for their co-operation. This part of the problem handling cycle ends with sending the final report to the initiator[79].

In chapter seven we gave some directions which a method for supporting defining a complex interdisciplinary societal problem should follow. These directions are, as far as possible, included in the method COMPRAM. COMPRAM stands for: COMplex PRoblem Analyzing Method. COMPRAM can be considered as a framework. COMPRAM assists the definition of a complex interdisciplinary societal problem, by going from an individual mental idea to a mutual conceptual model of the problem by an alternating process of individual preparation and group sessions. The iterative process of describing the problem in words, via defining the concepts, describing the phenomena and explaining the theoretical ideas on which the concepts and the phenomena are based, constructing a semantic model, a causal model and the system dynamic model will help defining the problem. The process of problem defining is guided by a facilitator and supported by several computer tools. Defining a problem this way takes considerable time, but we are convinced that with careful and thorough performance, the problem handling process will increase the chance of a more complete overview of the problem. This way of working will in the end save much time, effort and money, because a better analysis increases the chance of more effective changes to the problem.

[1] The method COMPRAM has been developed by DeTombe 1993.

[2] In Appendix II is described what a group support room should look like.

[3] External communication facilities are a telephone, a television, radio and a fax.

[4] See also chapter seven and Appendix II.

[5] The method COMPRAM covers the first sub-cycle of the problem handling process. In the (near) future the method will be extended so that it includes also the second sub-cycle of the problem handling process. However, this part is beyond the scope of this study.

[6] The method combines many information retrieval methods from the field of social science.

[7] Groupware is software for multi interactive parallel use. As far as we know Groupware has been, until now, mainly developed for brainstorming, ranking and multi-criteria analysis.

[8] A software tool developed at the University of Strathclyde by Collins & Ackerman .

[9] In some cases the facilitator can handle the software her/himself.

[10] This does not mean that the problem is less complicated, it could even be more complicated than before.

[11] See figure 10.

[12] We distinguish complex interdisciplinary societal problems along the lines of lack of knowledge, number of parties involved and urgency. The first category of complex interdisciplinary problems comprises knowledge problems where the lack of knowledge is the crux. The second category consists of problems in which different parties are involved with different and often contradictory interests: what we call a different party problem. This kind of problem is often combined with a knowledge problem. The third category of complex interdisciplinary societal problems we distinguish are urgent problems. Urgent problems combine a knowledge problem and a different party problem with time pressure.

[13] Although with some alterations it could also be applied to problems in which different parties are involved.

[14] Although it can also be applied in a mono-disciplinary team.

[15] Layer one up to six can be made in Hypertext, with programs such as Hypercard for the Apple operating system or Spinnaker Plus for a DOS operating system.

[16] This can be constructed on the computer with system dynamic software like Stella/Ithink for Apple and DOS, and the software package PowerSim for DOS.

[17] By describing the relation between the concepts and the phenomena we include also the relations within the concepts and within the phenomena. For the reader however, this is not repeated in the text all the time.

[18] The character T, H A, E and I, are used to indicate a theory (T), a hypotheses (H), an assumption (A), experience (E) or intuition (I). The numbers correspond with the definition of the concepts or the phenomena. The same indications are used in the descriptions of the other layers.

[19] Layer VII is not connected directly to the other layers, due to the limitations of the software. The first models are made with and connected by hypercard software. The last model is modeled by system dynamic software.

[20] See figure 10.

[21] In Appendix I there is a more extended description of the method COMPRAM.

[22] Aids stands for Acquired Immune Deficiency Syndrome.

[23] It will be clear that it is impossible to describe everything that will happen and all the knowledge that will result from the sessions. However, we may speculate about some effects of the problem handling process and illustrate some of the problem handling steps. Since it would be too detailed to illustrate each step, we have selected one or two examples in each session. Most of the examples are selected in such a way that the main events of the problem handling process are illustrated.

[24] As explained earlier, the method begins with the second phase of the first sub-cycle of the problem handling process.

[25] Each participant has a reference group.

[26] The virus that is active in Africa is called HIV-2.

[27] Among the first victims in the male homosexual population were a huge percentage of (often young, and very promising) artists. To name a few: Keith Haring, Andy Warhol, Freddy Mercury, Nurejev.

[28] In the USA the disease spreads rapidly now among heterosexual (colored) women of the poorer districts in the big cities, with terrible consequences for their children. One third of new-born children of infected women are also infected. Since many of these families are single (women) parent families, these children are orphaned at an early age. (WHO-EC/OMS-CE, 1992).

[29] Chin, J. (1990, p. 221) states:

"Factors that are responsible for global patterns of HIV infection/AIDS include: (1) time whan HIV entered or began to spread extensively in the population; and (2) the relative frequency of the three modes of HIV transmission - namely sexual, parental, and prenatal. Several broad though distinct epidemiological patterns of HIV infection have been decribed by WHO. In pattern I areas, the primary population groups affected are homosexual men and intravenous drug users; extensive spread of HIV began in the late 1970s/early 1980s. ......... This pattern is seen in North America, Western Europe, and Oceania. In pattern II areas, HIV infections/AIDS is found predominantly in sexually active heterosexuals; extensive spread of HIV probably began in the mid-to-late 1970s. ..............Pattern II areas are sub-Saharah, Africa and some parts of the Caribbean. Many Latin American countries (are) now classified as pattern I/II. Areas currently classified as pattern III include Asia, most Pacific countries (excluding Australia and New Zealand(Eastern Europe, North Africa, and the Middle-East. HIV was introduced into these areas in the early-to-mid 1980s. Although there is indigenous spread of the virus in most of these countries, the prevalence of both AIDS cases and HIV infections was low at the end of the 1980s with no clearly predominant mode of HIV transmission. However, the situation is changing rapidly in a few countries. During the late 1980s, the prevalence of HIV infections has greatly increased among intravenous drug users in Southeast Asia, especially in Thailand, where the prevalence is now nearly 50%; focal increases (up to 50%) have been recorded among female prostitutes in several cities in Thailand and India."

[30] In Asia and Africa the disease is spread mainly through heterosexual contact and blood transfusions.

[31] In Asia and Africa Aids is sometimes called a poverty disease, because women prostitutes who are forced to sell their bodies and sexual services to men out of poverty are getting infected.

[32] However, this idea is in contrast with the well-known knowledge about the many diseases for which there are no primary secondary and tertiary preventions, diseases such as some hart diseases, the flew, rheumatism, Alzheimer's disease.

[33] In 1918 influenza caused the death of many of people in Europe.

[34] However, in some cases of tuberculoses penicillin fails. After forty years, in which penicillin could fight the disease succesfully, tuberculoses is a fatal disease once more.

[35] See chapter nine.

[36] As described earlier (see chapter one) some aspects are studied by different disciplines, although, often based on other theoretical ideas and analyzed with different methods.

[37] HIV infected immigrants are not welcome in the USA.

[38] See for instance the opinion of the Roman Catholic church.

[39] See for instance the ego document 'Silverlake life'. Two homosexual men, partners, trying to cope with the disease, their relationship and with their family (Joslin & Massi, 1993).

[40] The value of the product of the problem handling process will depend on the selection of the team and on the way the selected experts can be regarded as experts in their field. Some of the persons that are interviewed in order to collect some general knowledge about the problem ( in P1-2) might give indications about who to select for this process. In order to cover some of the domains within the disciplines it is advisable to select the reference group but in a (slighly) different domain of the discipline in order to enlarge the knowledge of the team.

[41] This selection should only be regarded as an illustration not as a well-considered selection.

[42] HIV is a human retro virus (Stienstra, 1990, p. 437).

[43] The issue of what was later indicated as 'Aids' appeared in journals for the first time in 1981. Since that time the number of patients with Aids and the number of those infected with the Aids virus (officially called HIV), the so called seropositives, has increased dangerously (Stienstra, 1990, p. 437).

[44] Although it is not a straight linear decline (Dangerfield, 1990a, b).

[45] The strength of the immune system can be expressed by the number of T4-cells in the blood. With a normal intact immune system the number of T4-cells varies between about 1000 to 1500. When a person has less than 200 T4-cells, the person is supposed to have Aids. The T4-cells can decline to zero before death.

[46] A description of how the HIV attacks and how a person gets ill is given in Geelen (1990, p. 444).

[47] CDC stands for Centers for Disease Control in USA. See for the special role of CDC concerning Aids also Shilts (1987).

[48] GHI stands for Chief Medical Office of Health.

[49] See for a describtion of some of these desease Monette (1988)

[50] The characteristics of Aids seem to be different for different kinds of populations. For instance, karposi's sarcoma appears relatively often in homosexual males. (Shilts, 1987; Joslin & Massi, 1993).

[51] This means changing from sero-negative (not having the virus) to sero-positive (having the virus).

[52] There would seem to be theoretically other possibilities such as by saliva in the mouth, but there is no evidence that people have been infected this way.

[53] The definition of hetero and homo sexual is used here in the way people would address themselves.

[54] In 1985 a first generation of tests to detect HIV antibodies was developed (Akker, 1990). HIV positives can be tested by the HIV Elisa test (Stienstra, 1990, 438).

[55] See also Dangerfield, 1990a.

[56] Possibly having wounds due to (former) venereal diseases plays a role.

[57] The strategy of public education is to assume that from the moment of infection till the death of a person the virus can be transmitted. However there may be some time between the moment of sero-conversion and the moment one can infect another person. Research shows that the infection curve of the virus can be U shaped, with extremes at the beginning and the end of the period and a latent period in the middle (Dangerfield, 1990a,b).

[58] May 1994.

[59] Transfusion recipients 90 % confidence interval of the median 4.4 year and 9 year. Homosexuals 9,9 years ( 6,7 -13.5) For paediatric cases two populations first within 18 month 11 % of the children, the rest of the population: similar to adults 6.4 years (Public Health Service, 1990).

Tovo (1990, p. 290) "Mother to child transmission rate of infection can be estimated to be about 25% with a possible negative influence of breast feeding. Preterm delivery and low birthweight seem to be related to drug use during pregnancy, not to intra uterine infection."

[60] See for an overview of research on interventions Dutch Program Committee for Aids Research, 1992, pp. 23-24.

[61] See primary, secondary and tertiary prevention in chapter three.

[62] This played an important role in the conflict between Montagnier, a researcher of the virus from France, and De Gallo, a researcher of the virus from the USA. The conflict was over who discovered the virus first. This way the discovery of the virus could be connected to a particular sample. (Tracking the Virus, video).

[63] AZT stands for zidovudine or Retrovir (name of the firm).

[64] AZT is a medicine that slows down the declining of the T-cells. It also gives in many cases a temporary relieve of the diseases and in this way increases the quality of life. However some patients are allergic to it. Those patience will get DDI , which has more or less the same effect as AZT. However both medicines give rise to different but unpleasant additional side effects. The medicines as such is rather toxic. Another negative point is that the patient will become resistant to the medicines, so that after a while the medicines looses its efficacy. There are now experiments that alternate both medicines in such a way that the patient can profit longer from the treatment and suffers fewer negative aspects. However this treatment also looses its effect in the end.

[65] See for information about T-cells Gruters, Terpstra, Lange, Roos, Harkema, Mulder. de Wolf, Schellekens & Miedema (1991) and Dutch Program Committee for Aids Research (1992, p. 62-64). CD4 T-cell function in HIV infection. The declining of the number of T-cells indicate a decline of the immune system.

[66] By 'controlled ' we mean that both partners are sero negative.

[67] Knowing something rationally, that must be done, is different from actually doing it. Most people know what is not good for them, yet they often do it, mainly because there is some immediate benefit and because people underestimate or trivialize the negative consequences, or have a fatalistic approach. We believe that in order to use a condom more is needed than mere cognitive knowledge. It is also a matter of having money to buy them, of daring to buy them, daring to have them with you, being willing to use them. A person should be able to negotiate with the partner about using it and last but not least be able to use the condom properly in practice. See also risk factors for transmission of HIV and changes in sexual behavior (Dutch Program Committee for Aids Research, 1992, p. 13).

[68] See also section 9. 1 hidden agendas.

[69] Research is not quite clear about the moment of sero-conversion.

[70] See congress: Women and AIDS in December 1990, which was held in several European countries including the Netherlands.

[71] Some fundamentalists are against vaccination, among them Jehovah's Witnesses. They think that this is interfering with the plans of God. An example of this kind of behavior is the small polio outbreak at the end of 1992 in The Netherlands among the fundamentalists, who were on principle not vaccinated.

[72] The differences in power between participants, the danger of group think and the extent to which this can be prevented are discussed in chapter seven.

[73] Session seven can contain several sessions needed for repeating the activities in the former sessions now according to the desired scope and aggregation level.

[74] In sexual relationships one can divide people into different populations. For instance, the homosexual population, the prostitute's clients, the bi-sexual population, and the heterosexual population. Most people select their partners in one population or two populations. Some people select partners from more than two populations. The chance of getting infected by a non-protected sexual contact differs from one population to another, depending on the rate of infection within the population.

[75] In 1993 80% of the death of homosexual men in Amsterdam was caused by AIDS.

[76] See chapter three.

[77] Different domain experts are invited.

[78] Based on these reports further steps towards the handling of the problem can be taken.

[79] The initiator is in many cases not directly involved in the problem handling process. In practice many problems are handled within the time and more or less according to the way initiators want this. However, initiators do not always have the benefit for all in mind when they give the assignment, but even if they have, they are limited themselves in the way they view the problem. However, for such an important problem as is reflected here, as well as for many complex interdisciplinary societal problems which have a large impact on society, the facilitator and the problem handling team should have the freedom (including time and money) to really analyze the problem thoroughly. The facilitator and the team should be allowed to give a critical view on all the things that have been done and should be done regarding the problem. The facilitator should have, what you may call, an ' academic freedom'.