Cict.demon.co.uk

DRAFT TEXT to be published in Organisational Systems: International Perspectives on Research and Development, Berkeley, D. & Dikaiou, M. (Eds), Innovating social support. Examples and designs
Centre for Innovation and Cooperative Technology1 Organisations differ in many ways. Two extremes are especially interesting. Someorganisations are tyrannical. Their members are committed to strict rules, activities outsidethe organisation are not considered of relevance to the organisation. Others are freedomloving. Little effort is needed to keep to the rules. Members feel supported. Their locus ofcontrol is close, they feel appreciated and honoured. They are committed to contribute toand maintain the organisation.
Neither kind of organisation seems ineffective. Both seem prone to dangers – to fragment in the case of the second type of organisation, to become too strict andimprisoning in the case of the first. In Western society it is impossible to doubt a preferencefor the second type. Interestingly, this preference still proves difficult to implement,especially in the area of social helping. There still is a tendency to imprison, and to tyrannisethe helpee by the values and norms of the helper.
This defines the problem discussed in this paper: what may help organisations similar to the first type to change and become similar to the second? It is proposed that the changecan be effected by changing the communication structure. It should allow members toaddress each other to provide precisely what is needed to act as members of a stablecollective or organisation.
In the paper examples are presented of how to design the research needed to identify proper communication structure. A general pattern is identified. Proper structures are thosethat allow members to add to their existing competences as members of stable collectives.
Stability, however, is a relative notion. It refers to self-organising new competences as a nextstep – after having achieved previous competences. The search for such stabilityimplements a long needed research program.
Introduction
A few years ago a Dutch social worker reported an interesting experience. Aclient requested regular appointments to help avoid being battered by herhusband. A series of helping sessions was offered. This she refused, as sheclaimed only to need the appearance of an appointment to keep her husband inline. Only later did the social worker accept this procedure, which, at first, sheexperienced as professional failure.
This anecdote illustrates what appears to be a general dilemma. On the one hand there is client behaviour that fits expectations. Social workers usuallyclaim to be in favour of clients solving their own problems, precisely as theclient in the anecdote intended. On the other hand there is client behaviour that Affiliated with the Lincoln Research Centre, Michaelgate 20, Lincoln LN1 3BT.
does not fit expectations. Social workers do not appear to be prepared toconsider themselves a collective resource.
This dilemma obviously derives from a stereotype, that is from the way social workers see themselves – as individuals whose competence rests onpersonal initiative, experience and the results of research. The lady requestingappointments had a different view. She emphasised the role of the collective ofprofessionals. To her the profession itself was a resource, the competence ofwhich to her overshadowed that of individual workers.
It is seductive to reject the latter view out of hand – but also to take it seriously. Social workers may be trapped in their own definitions, bothpractically and research-wise. If so the anecdote suggests an interesting way toget out of the traps and improve social help. Clients want to contributethemselves. Helping thus constitutes a wider problem than just improving theskills of the individual social worker - be it by research or not.
In this paper I aim to explore this wider problem. It is important – as may be gleaned from the fact that similar anecdotes are found in many areas of sociallife. Whom should the traveller criticise, the conductor or the rail company? Italso appears under-explored. It is possible to think of other models for helpingthan those in which the basis of the helper’s competence is restricted to helpers’experience.
Elements of the problem
Problematisation usually implies picking a smaller problem and stretching it soit starts to explain wider problems. Interestingly, in this process the choice ofthe smaller problem does not appear to matter – as long as its definition derivesfrom outside the area where the problem is to be diagnosed. For example, itmay be someone claiming that a watch is not working, and hence wishing tohelp and ‘cure’ the defective parts ‘for ever after’.
As a first step in stretching this (smaller) problem one may assume that it is defined by the watch itself. Or more practically, by the human being who is tobe helped and who may wish to define his or her own defect. The helpee mayhave a wish other than to ‘work’ again, for example to work differently or tofunction better. The only help needed in this case is to ensure the ‘for ever after’,that is the stability2 of what is achieved.
This case appears similar to that of the lady asking for appointments. She did not want her problem to be defined and solved by the social worker. Sheclearly considered her approach (too) temporary. What the lady lacked wassupport to achieve stability, something the social worker found difficult to give.
Formal appointments with her organisation would be sufficient, apparently, toprovide the necessary anchor.
It seems natural to stretch the problem of helping even further, to where the latter aim, anchoring achievement, is the helpee’s as well. The term we will usefor this combination is self-organisation. It refers to the helpee’s realisation of twoaims in tandem: solving a special problem or achieving a special effect as wellas maintaining that solution or that effect. The help required now is to ‘stabilisestabilisation’.
The next step seems obvious again – to further unfold and problematise the wider problem by stretching the notion of helping to ‘stabilising stabilisation ofstabilisation’, etc. Two possibilities may be envisaged. It may be that eventually Being stable is not the same as being in equilibrium: it requires an effort, and hence is morelike being (far) from equilibrium.
a final stretching is identified, e.g. helping to stabilise against any disturbance(Carse, 1987). The other alternative refers to what one expects in practice – thateverybody needs help at some level3.
In the second case the notion of competence appears useful. It is defined as confining change, for example increases in skill, by bestowing a qualification –or a set of constraints which a person is able to stabilise. The client requesting aregular appointment thus can be said to have identified the last level where shedidn’t need help to stabilise the qualification that would make her competent toavoid battering, or the first where she did.
Problematising the notion of help confirms the need for something beyond the traditional approach, that is diagnosing the problem of the helpee andsearching for its solution. What is needed goes far beyond knowing what thehelper may do irrespective of the helpee. It reverses the roles of the helper andthe helpee. To get help the latter must be able to challenge and bring the helperto help only at the level requested.
The change from paying attention mainly to what the helper may do, as is stillcommon among professionals, to paying attention to what the helpee maychoose is quite fundamental. It raises two questions. The first is what it impliesfor the build-up of competence, and for the role of research. The second is whatit looks like in familiar terms, for example of developments in practical areas ofhelping. This will be discussed now.
One major and well-known area is helping youngsters to grow up without too much violence and avoiding criminalisation. It has become clear that beingstrict or applying zero tolerance does not help much, or negatively. This iseasily explained. It is not sufficient to tell youngsters what to do, that is ‘replace’what they want to do by what their parents and teachers want them to do ‘forever after’ (Van der Doef, 1992).
What adolescents appear to aim for is to choose their own achievements, as well as their own ways of stabilising these. Most social actors appear well awareof this, but few seem willing to provide the necessary space. This is especiallythe case where it concerns sexual behaviour. Maintaining ‘proper’ behaviouroften is thought to require forceful control and even ‘policing’. This approach isdoomed to lead to resistance.
A symptom of such resistance is teenage pregnancy, which is high in countries like Peru, but also high in the United Kingdom4. Becoming pregnantmay be an accident, but it also appears due to a desire to step out of the child’sposition and create a family oneself – as a way to stabilise an adult role. Thisinterpretation is supported by a project in Peru, where this kind of stabilisationwas replaced by that via peer groups (Ramella, 2001).
Adolescents were invited to participate in ‘grupos de adolescentes’, to create their own culture and earn their own money (for example by organisingparties). This was helped by providing each group with a camera and askingthem to tell about their community and about their interaction as boys and girls.
Some of their stories told how the groups helped to provide stability to theyoungsters, in case their families did not.
A person who is able to act stabilised against all obstacles obviously appears ‘strong’, or astrong actor. Similarly, those who achieve a new level of stability, beyond the last one,become ‘stronger’ actors.
Teenage pregnancy are much lower in countries like the Netherlands (in the latter case 10times). This seems to be due to a strategy of providing early education (from the age of 4onward), and of providing alternative projects (like building 17th century ships).
One of the results was that the frequency of teenage pregnancy in the groups dropped considerably compared to elsewhere. Possibly even more importantwas that the groups became autonomous agents. They were listened to as ableto provide help to the community. Interestingly, members of the groups, whengrowing older, still appeared more independent, more confident of their owncompetence than others of their age.
There are other areas where help frequently has been (and often is) modelled onreplacing or repairing ‘parts’ and hence on imposing what is considered anachievement. The area of social work is particularly well known in this respect.
One aspect that received much attention is people becoming dependent on helpand hence no longer self-organising. This has been called the ‘sucking’ power ofthe helping society (De Gier, 2001; p. 27).
This process is characterised by increasing numbers of people seeking help but also of social helpers – far beyond what one would expect based onexperience. In the Netherlands this has led to abrupt changes in social policy5,usually argued to be due to economic developments, in particular to a saggingeconomy and rising costs. It was expected that costs would decrease whenpeople would be called upon to self-organise (De Gier, 2001, p. 26).
Another interpretation is a loss of tradition. In former times a strong sense of community could be assumed, with clear norms and values, without muchneed for ‘policing’. When this sense diminished, social workers and others wereasked to provide new concepts. A series of concepts resulted justifying reducedsocial support, for example social integration, social cohesion, social capital,civil society, etc. (De Gier, 2001, p. 26).
Sen (1999; see also De Gier, o.c., p. 37) interprets social ills, such as poverty, as due to something like our ‘watch problem’. People should become self-organising agents – and develop their ‘capabilities’, or in our terminology, theircompetences. The stabilising anchors are education, health care and legal andsocial security. People need to care for themselves as well as for theirenvironments (see also Bakker, 1987).
An increased emphasis on self-organisation can also be found in areas such as management. The prime example is human resource management (HRM). Itopposes the Taylorian approach which is strongly reminiscent of our ‘watchproblem’ and on replacing ‘defective’ parts. HRM emphasises mutual respect –and insists on self-organisation as a way to contribute to the viability oforganisations (Van Gent and Van der Zee, 2001; Chia, 1996).
Researching self-organisation
There appear to be good reasons to re-interpret the notion of helping, therefore.
The anecdote gave a first indication of what this might look like. Next it provedpossible to problematise the traditional notion and to develop one that evenseems its reverse. Developments in various social areas demonstrated atendency in the same direction. What is needed now are research designs toimplement this reversed notion.
To develop such designs, the main step is distinguishing between what needs to be reversed, and what does not. Following the re-interpretation, forexample, one should search for what will help any helpee, irrespective of the This includes strongly decreasing support for a discipline called andragology (whichincluded the study of social work, of social helping, of the effects of the built environment,etc.) that in its heyday attracted some 400 new students each year.
helper. This clearly is very similar to the traditional requirement of professionalknowledge, that it be of use to any helper, irrespective of the helpee’s problem.
Besides such similarities, there are differences.
The major instance is that helpees will push help to just beyond the point where they don’t need it. One may think here of a complaint most of us hadwhen parents tried to help with our homework. They tended to explain toomuch. Helpees also must challenge helpers not to help too much – unlike in thetraditional case which required helpers to ‘know’, or observe and diagnosewithout interference of helpees.
This difference as well as the similarity may be combined in a model from which the desired research designs are generated. First, one may note that self-organisation appears a special form of the traditional approach. The latter alsoadds something, or in traditional terms, adds something that stabilises byexhausting all observational variety. Here the emphasis is on discovering whatexhausts ‘for ever after’.
Second, some form of communication is needed to challenge helpers at the right level (just beyond the point where no help is needed). Suchcommunication should have a precise structure, therefore: it should hide whatis not needed and make explicit what is needed. This obviously will require thatthere is feedback between helpers and helpees, meaning that the structure linksboth with sufficient speed.
The model from which the designs are to be derived should include, therefore, identification of those structures that allow helpers and helpees toconstitute a collective that, as a whole, exhausts all observational variety. Thiscollective should be stable against any (observable) activities that helpers (orothers) may impose on helpees6. The task is to discover the structures thatenable such collectives to develop.
It may be thought that this task is very difficult or even impossible. This obviously does not mean that it has not been attempted. An example in the areaof management is the ‘methodology’ proposed by Checkland and Scholes(1990). It aims to help (all or some) members of an organisation to communicateand become a collective that is competent to identify and solve (theorganisation’s) problems.
This methodology aims to make the collective stable by eliciting and including as much observational variety as possible – initially when membersget together as well as later when solutions to problems are checked againstexisting experience. This approach is claimed to, indeed, address members ofthe organisation just beyond where they (still) are able to ‘manage’, and therebyturn them into helpers.
It also shows some drawbacks, which seem to make it only partially count as a research design. It does not explicitly provide to create an exhaustivecollective. For example, its checks for stability do not include observationalvariety from outside the organisation. This makes it difficult to stabilise ‘forever after’7. Also, it does not identify whether structures other than the oneused, contribute to addressing helpers on the desired level.
Designs in social helping
Various lines of research have developed that nowadays attempt to identify what helpscollectives develop and maintain themselves (Briggs and Peat, 1985; Waldrop, 1992; Lewin,1992; Casti, 1995; Penrose, 1995).
Convergence to such a result is not guaranteed by repetition of the methodology, althoughthis is claimed.
Although ‘helping methodologies’ are widely available in management8, this isless so in the area of social work, adolescent and adult education and the designof the built environment to help people achieve a high(er) quality of life. This isnot to say that no such methodologies or research designs have been developed(e.g. Denzin, 1997; Erlandson et al., 1993; Guba, 1990), only that none isdominant.
There have been many attempts to design and test suitable designs. An example is a study by Nieborg (2000). She was concerned about the fact that thedivision of labour in Dutch households has not changed much, even though thenumber of women participating in the Dutch labour market did risespectacularly over the past two decades. Women still are supposed to providemost of the childcare, for example.
This clearly is not due to a lack of effort, or even awareness. For some time the Dutch government created crèches and other forms of support at the rate ofmore than a quarter of a billion US dollars a year – without much effect on thedivision of labour. This suggests that the model of helping was inadequate. Justlike social work in the 1970s, it sucked up extensive resources, producing littlepositive, and often quite negative effects.
The author decided to explore a different model. She interviewed members of families who did change their division of labour as well as of those whodidn’t, and also marriage counsellors, government ministers and employers.
This material was analysed as to the structures families used to function asstable collectives, or as strong actors. It was shown that not changing the divisionof labour did help families to be ‘strong’.
Evidence was collected that when families tried to change the division of labour, they became ‘weak(er)’ actors compared to actors in business andgovernment. Eventually the author was able to formulate statements of advicethat members of families could use to ‘think about’ their roles. This helped themmanage and better resist stereotypes such as ‘managers have to socialise afterhours at cocktail parties’.
Vahl (1994) was invited to help decide whether to continue experimental serviceteams. Conventional evaluations would have provided exhaustive descriptionsof what was achieved, which then would be judged in terms of the originalobjectives. However, this procedure usually leaves undefined what precisely isto be described, and hence allows external influences on the selection, e.g. thecommissioner’s.
To avoid this possibility, the author decided to first help members to self- organise the teams as collectives so they could stabilise against being influencedby external requirements. This would allow the teams to be judged withoutsuch influence. She then selected a communication structure that might havethis effect, a hypothesis she eventually could test successfully by developing theteams.
The author used a structure derived from the work of Axelrod (1984) and Howard (1971). It suggests communication to consist of sentences of the form ‘Ifobservation A (of the action of a previous actor), (let the next actor) do action X’.
Members of the team identified whether (and which ) sentences in their dailycommunication showed this form. This led to full descriptions of existinginteractions among the team members.
There are many other ‘methodologies’ with similar aims (e.g. Stafford Beer, 1985; Kompierand Marcelissen, 1990).
These interactions obviously had not been sufficient for stabilisation (otherwise conventional designs could have been applied). The communicationstructure needed to be improved, therefore. It was accepted that team memberswould do this themselves. They were taught to update their sentences daily, e.g.
to change X when A would change (such as new kinds of clients beingadmitted), or A when X would change, or both.
This updating procedure proved sufficient to help the teams self-organise and become coherent, resilient and strong as actors. Feelings of enthusiasmreturned, costs were reduced and the quality of the service to clients increasedand eventually proved maximised given the resources available. The decisionwhether to continue could be based on (maximum) possible results, rather thanon results as achieved.
Van Geen (1989) wished to improve on the life of residents in homes for theelderly. She decided to identify a communication structure by using availableforms of statistical analyses. Statements from residents about what it means tolive in such homes could be reduced to a set of seven concepts. They deal withfriendship and freedom, food, organisation, comfort, belongingness and staffbehaviour.
The structure consisting of these concepts was made available to residents’ committees. It thereby became part of a daily updating procedure. Residentsproved better able to address helpers so their control over their own livesincreased. This led to the introduction of the same structure in another 150homes. Evaluations indicate that residents started to take new initiatives. Theirhealth improved9.
Similar effects were noted by Van Haaster (1991) who explored a structure to strengthen individual identity against staff control in a Day Activity Centre forex-psychiatric patients. Staff learnt not to behave as controllers. Eventually mostforms of violent behaviour disappeared. His procedure again consisted ofhelping patients to identify and daily update a structure that helped to maintain(life in) the Centre.
The structures that appear useful in helping apparently thus vary quite a bit.
The same holds for the daily updating as Van den Berge et al (1980) show. Theywanted to support homosexual men running the risk of being beaten (‘gaybashing’). What the authors want to avoid was that the latter would be unableto visit certain places or would have to learn, for example, karate. Suchrestrictions imply a (too) high social investment.
The authors argued that being beaten requires a ‘qualification’ by the attackers as ‘victim’. This induces a ‘competence’ in being bashed – just ashelpees become qualified and competent in being ‘dependent’. It turned outthat there were anecdotes about how to address attackers successfully, to avoidthe qualification. The authors collected such stories and improved them oncriteria such as ‘ease of understanding’, ‘grippingness’, etc.
One of the anecdotes, for example, tells how someone is walking in the park.
When he attracts the attention of potential bashers, he starts to flap his arms andyell ‘Peeweet, peeweet’, as if he is a bird. His assailants walk away in disgust.
Help in the form of providing what stabilises competence beyond existing levels does notleave the individual unchanged. It requires that he or she accepts the constraints involved,and hence implies changes in norms and values. For example, one can not be member of anexhaustive collective if one does not intend to behave honestly, that is on the basis of hidden(and hence non-exhausted) purposes.
The new form of addressing had changed the qualification of ‘victim’ to ‘madperson’, or to someone of no interest to gay bashers. This change is cheap. Itrequires no physical training (but good timing).
A selection of such anecdotes (with an explanation on how to use them) was published and about 1500 copies were sold in Amsterdam. Interviews withsome of the buyers indicated that the anecdotes were used as intended. Theywere not interpreted as (repeatable) solutions to a particular problem or asgeneral strategies. Rather, they introduced a new ‘way of thinking’, and helpedto create new collectives, with new qualifications.
What seems to have contributed most to this result was the power of the anecdotes to suggest forms of addressing. A dangerous situation is described,then behaviour that induces a change in the qualification of the subject of theanecdote, then the surprising consequences of that change in terms of how otherpeople start to behave differently. Anecdotes are quite near to the ‘moral tales’that most cultures cherish.
After a few years, the anecdotes lost their attractiveness – at a rate similar tothe decrease in the overall number of beatings. This presumably was due toother factors, however, for example the legalisation of homosexual households.
However, what is important in the context of this paper is the structure that wasintroduced (anecdotes) and the way people proved able to use it to update theirability to cope with (group) aggression.
Comments
The above examples were intended to clarify three kinds of issues. Firstly, theissue of deciding whether to introduce a pre-selected communication structure,or to design a new one. Secondly, the issue of how to test whether the resultingcollective (if any) is sufficiently stable. Thirdly, the issue of spending (andpossibly minimising) the effort of daily updating and testing the structure toensure that the collective is maintained.
Although the examples already diverge substantially on the first two issues from the usual approaches to improving helping, it is especially the third issuethat makes the difference. In traditional approaches, once something has provedits mettle, it is treated as requiring no further maintenance. This is not the casefor the communication structures that helpees need to address helpers at someproper level.
Behind this difference is the idea of stability. What makes for self- organisation? A practical example may help to answer this question. One maythink of a support team for the disabled, or of a shop wishing to sell tea. Toimprove on them two strategies may be considered. The first is to getknowledge about clients or customers. The second to provide information abouthow the organisations will serve.
The first strategy assumes that stability is achieved when variation in the needs of all customers or clients can be identified and proved bounded. Noattempt of this kind has succeeded, unfortunately, presumably as such needsoften change. This is where the second strategy comes in. The am is to design astable and informative ‘shop window’. It allows restricting and boundingvariations in the needs of available customers and clients10.
It usually is taken for granted that no further effort is needed to maintain the bounds on the variation in clients’ needs – ‘for ever after’ – and hence on the 10 In other words, although no shop may ever ‘know’ its customers, it will be able to self- organise itself to serve its customers.
knowledge needed to offer what is attractive. The second strategy requires thatthe display (and its limitations) be updated frequently, or even continuously –on penalty of becoming less attractive and of losing clients. Changes in needsand tastes must be adapted to.
There is thus a basic difference between the two strategies. In the second strategy helping implies keeping track of the next level where helpees needsupport and stabilisation. This implies a form of (mutual) communication, andhence some (possibly changing) structure to address helpers. In other words,what is stable now does not have to stay the same: it may change wheneverhelpees develop.
We may refer to this form of stabilisation as a criterion, the criterion that makes helping in the sense of addressing helpers an instance of research11.
Without the criterion, there would only be a project, lacking systematiccomparison of later achievements with previous ones. The criterion ensures thataccess to communication structures will be general, or rather freely accessible toall, and in that sense, democratic.
The counterpart to spending effort to maintain stability is that it can be destroyed. As an extreme, one may think of an asteroid wiping out theorganisation’s clients, or a war destroying the shop. If stability is to bemaintained in such situations, it obviously will be at (too) great cost. In theexample of the experimental teams (Vahl, 1994), nothing else was intended butto stabilise against the influence of the commissioner.
Another interpretation of stability is that one is prepared to spend effort to deal with changes that threaten stability12. The latter obviously may includeactions by clients – similar to those of commissioners. In other words, effectivecommunication structures must lead to collectives being selective. Acceptingmembership implies that members commit themselves to respond as much asthey are capable of when addressed.
Conclusion
Starting from an anecdote it was argued in this paper that helping often isinterpreted ineffectively as a role only for the helper. He or she is expected toact purposefully by diagnosing ‘definitively’ what helpees need. The anecdotesuggests a change in emphasis. It should be the helpee who addresses the levelfrom where help is provided, using a communication structure that does notimpair competences below that level.
As a next step, it was attempted to exemplify research designs that help helpees to use proper structures. This involves identifying what structuresallow for the development of collectives that are sufficiently stable to serve asanchors for helpees to self-organise. A number of examples were described toclarify the notion of communication structure as well as of regular updating(and hence testing).
It was argued that an interpretation of helping as helping to self-organise fits in with recent developments, both in society and in the area of helping itself.
This makes it even more important to develop research designs to implementthis interpretation. The difficulty such designs face is that results depend on an 11 Traditionally this criterion does not refer to stability, but rather to timelessness, to placelesness and to being usable by anybody.
12 It is surmised that the failure of Checkland’s ‘methodology’ as a research design is, especially, that it does not anticipate spending effort in this way. See also Kaufmann (2000).
unusual notion of stability – a dynamic one in that it requires effort to anticipatespending effort on stabilisation.
References
Axelrod, R. (1984), The evolution of co-operation. New York: Basic Books Bakker, B. (1987), Leidraden voor zelfhulp. Amsterdam: Ph.D. Thesis Universiteit van AmsterdamBeer, S. (1985), Diagnosing the system for organisations. Chichester: WileyBerge, F. van den, W. Bossewinkel, S. Groeneveld, M. Muis, H. Wildschut, Flikkers en agressie (1980), Amsterdam: Scriptie Universiteit vanAmsterdam Briggs, J.P. and Peat, F.D. (1985), Looking glass universe. The emerging science of wholeness. Glasgow: Fontana Paperbacks Carse, J.P. (1987), Finite and Infinite Games. Harmondsworth, Middlesex: Casti, J. (1995). Complexification , Explaining A Paradoxical World Through The Science Of Surprise. New York: Harper Perennial Checkland, P. and Scholes, J. (1990), Soft systems methodology in action.
Chia, R. (1996), Organizational analysis as deconstructive praxis. Berlin: De GruyterDenzin, N.K. (1997), Interpretive ethnography: Ethnographic practices for the 21st Doef, P. van der, ed. (1992), Interdisciplinaritteit in de jeugdhulpverlening en de adolescentenzorg. Amersfoort: AccoErlandson, D.A., E.L. Harris, B.L. Skipper, S.D. Allen (1993), Doing naturalistic inquiry. A guide to methods. London: Sage Gent, B. van & Zee, H. van der, eds. (2001), Handboek van human resource Geen, V. van, (1989), Meer greep op eigen leven. Amsterdam: Zwets & ZeitlingerGier, E. de (2001), De sociologische interventie. Wetenschap en sociaal beleid.
Guba, E.G. (1990), The alternative paradigm dialogue. In: E.G. Guba (ed.), The paradigm dialogue. Newbury Park: Sage, p. 17-30 Haaster, H. van (1991), Wartaal. Amsterdam: Thesis PublishersHoward, N. (1971), Paradoxes of rationality: theory of metagames and political beha- Kaufmann, S. (2000), Investigations. Oxford: University PressKompier, M.A.J., & F.H.G. Marcelissen (1990), Handboek Werkstress. Amsterdam: Lewin, R. (1992), Complexity. Life at the edge of chaos. New York: MacMillanNieborg, S. (2000). Gedeelde zorg: gedeeld werk. Onderzoek naar verandering van de arbeidsverdeling in het gezin. Amsterdam: Ph.D. thesis Universiteitvan Amsterdam Penrose, R. (1995), Shadows of the mind. A search for the missing science of Ramella, M. (2001), Taking part: a study of adolescent sexual health promotion in Peru. London: Ph.D. thesis London School of Economics Sen, A. (1999), Development as Freedom. Oxford: Oxford University PressVahl, M. (1994), Improving mental health services in Calderdale. Report of the Centre for Systems Studies. Hull: University of Hull Waldrop, M. (1992), Complexity. New York: Simon and Schuster Zeeuw, G. de (2001), Three phases of science: a methodological exploration.

Source: http://www.cict.demon.co.uk/Innovating02.pdf