STHV special issue post-normal science

If post-normal science is the solution, what isthe problem?

The politics of activist environmental science

REVISED submitted

Anna Wesselink1

Sustainability Research Institute

School of Earth and Environment

University of Leeds

United Kingdom

Rob Hoppe

Department of Science, Technology, Health and Policy Studies

School of Management and Governance

University of Twente

The Netherlands

Abstract

Post-normal science (PNS) is presented by its proponents as anew way of doing science that deals with uncertainties, valuediversity or antagonism, and high decision stakes and urgency,with the ultimate goal of remedying the pathologies of the

1 Corresponding author a.wesselink@see.leeds.ac.uk

1

global industrial system for which, according to Funtowicz andRavetz (1993, 739) existing science forms the basis. Wecritically examine whether PNS can fulfil this claim in thelight of empirical and theoretical work on politics andpolicymaking. We credit PNS as innovative frontrunner inraising important issues regarding the limited problem-solvingcapacity of ‘normal science’ and ‘professional consultancy’.Yet, we notice that PNS lacks important considerations aboutthe governance of problems and aspects of participatory anddeliberative democracy. PNS in effect implies thatmethodological ‘ratiocination’ would prevail over politicaldeliberation and democratic interaction, and that merelychanging scientific input in public policymaking would havethe power to change its outcomes. This scientistic hubris canbe traced back to PNS’ origin in concerned scientists’activism, which in effect accessed the political arena throughthe scientific entrance. We conclude that the art of politicsneeds to come back into the discussion on environmentalproblems if societal change is to occur.

Keywords: post-normal science, science-policy interface,boundary work, frame reflection, policymaking, power, problemtypology

1 What we are going to do

In this paper we will analyse the potential of PNS to fulfilits own objectives from a political and policy scienceperspective. We argue that presenting PNS as ‘science’, whileat the same time advocating changes to policy makingprocesses, was conceptually fuzzy and unhelpful. Gettingthings done in the policy arena involves more than providingthe right science: it involves doing politics, i.e. usingpower and influence strategically. To all intent and purpose,with their proposal for PNS Funtowicz and Ravetz engage in

2

politics with a political agenda that can be described as‘green’ and ‘democratising’. ‘Green’ because they want PNS tohelp solve environmental problems, ‘democratising’ becausethey want to include actors whose (environmental) values areat present underrepresented in designing solutions. This isindeed a political agenda and not only a normative choice fora particular type of research. Policy issues of risks andenvironmental concerns are resolved in the political arena,and it is there that decisions are made who can participate indecision making, including the designing of solutions (usingPNS or normal science or professional consultancy). To discussthis political agenda and its consequences on theeffectiveness of PNS is the theme of this paper.

In Section 2 we show how the PNS proposal was steeped inpolitics although presented as a new way of doing science.This contradiction helps to explain why PNS has beenimplemented either as a scientific method for makinguncertainty explicit, or as a sensitising concept (Section 3).We suggest that reframing PNS as a proposal for policyanalysis with deliberation and governance as separate elementsclarifies how PNS has fared since its launch (Section 4). Webuild up an understanding of the governance aspects of such aPNS-type policy analysis from policy science literature, alongthe way explaining why PNS ran into resistance at severalgovernance levels. We thereby aim to contribute to cross-fertilisation between two bodies of literature: STS and policyscience.

2 The political agenda of post-normal science

In their well-known 1993 paper Funtowicz and Ravetz state theobjective of post-normal science (PNS): ‘science is now calledon to remedy the pathologies of the global industrial systemof which it forms the basis’ (Funtowicz and Ravetz 1993, 739).

3

This new, PNS method for doing science was tasked with‘cleanup and survival’ (Ravetz and Funtowicz 1999, 641), to bean ‘emerging science’ that can deal with ‘the environmentalcrises of this century’ (Ravetz 2006, 275). In case of thehigh systems’ uncertainties, intense value contestation, highdecision stakes and political urgency in environmentalproblems they recommend doing PNS because ‘the puzzle-solvingexercises of normal science [..] are no longer appropriate forthe resolution of policy issues of risks and the environment’(Funtowicz and Ravetz 1993, 750). In contrast, in PNS‘uncertainty is not banished but is managed, and values arenot presupposed but are made explicit’ (Funtowicz and Ravetz1993, 740).

Underlying their framing is a distrust in scientists’willingness and ability not to speak only for commercial orregulatory interests, which is one reason why they considerthat traditional scientific peer-review is not enough toguarantee provision of information. Ravetz (2003) asserts that‘[t]he attempts at an ‘experts’ monopoly’ of the management ofrisk problems failed signally’, mainstream science isreductionist in style and increasingly linked to industry,while ‘regulators try to control [the risks] [..] whileallowing technology to proceed’ (Ravetz 2003, 812). Funtowiczand Ravetz’ strategy to counteract this experts’ monopoly isto extend the peer review, for ‘[o]nly a dialogue between allsides, in which scientific expertise takes its place at thetable with local and environmental concerns, can achievecreative solutions to such problems [..]. Otherwise, eithercrude commercial pressures, inept bureaucratic regulations, orcounterproductive protests will dominate [..]’ (Funtowicz andRavetz 1993, 751). They propose ‘the inclusion of [..]legitimate participants [..] For example, persons directlyaffected by an environmental problem [..] perform a functionanalogous to that of professional colleagues in the peer-review or refereeing process in traditional science [..]’

4

(Funtowicz and Ravetz 1993, 752). These formulations indicatea commitment to a political agenda which is ‘green’ as well as‘democratising’. While the latter can justifiably be apolitical objective on its own right, Funtowicz and Ravetzpresent it as an instrumental precondition for achievingbetter environmental management.

In their 1993 paper Funtowicz and Ravetz thereby display agreat faith in science’s ability to bring about changes insociety’s dealing with environmental problems. They imply thatthese problems will be solved if the right kind of science isbeing done, and vice versa: that modifications to scientificpractices are essential to solving the environmental crisis.They are careful to stress that PNS is a way of doing science.For example, when discussing the involvement of extended peercommunities, Funtowicz and Ravetz assert that this serves toimprove the quality of the science, not as a way to getinvolved in political decision making. They explicitly limitthe involvement of extended peer communities to ‘debates onquality’ (Ravetz and Funtowicz 1993, 752), although ‘[o]noccasion, the legitimate work of extended peer communities caneven go beyond the reactive tasks of quality assessment andpolicy debate [and] involve concerned citizens doing thedisciplined research’ (Ravetz and Funtowicz 1993, 752 ouremphasis). In PNS ‘[t]he problems are set and the solutionsevaluated by the criteria of the broader communities. Thuspost-normal science is indeed a type of science, and notmerely politics or public participation’ (Funtowicz and Ravetz1993, 751). At the same time they imply that the inclusion ofextended peer communities ensures dealing with values andstakes: ‘values are not presupposed but are made explicit’(Funtowicz and Ravetz 1993, 741) and ‘all those with a stakein the dialogue on the issue’ are invited to participate inpeer review (Funtowicz and Ravetz 1993, 739). How ‘merelypolitics or public participation’ and dealing with values andstakes differ from framing and putting problems on decision

5

agendas, and evaluating and adopting solutions on behalf ofbroader communities they do not explain: this may be exactlywhat politics and public participation are all about.

While the 1993 presentation of PNS stressed its scientificcharacter, in his 2006 reassessment of the purpose and successof PNS Ravetz accepts that ‘PNS has always had strongpolitical aspects’ (Ravetz 2006, 276). In his view, involvingan extended peer community is at the core of this politicalcommitment (Ravetz 2006, 275)2, through which ‘PNS can help toshape a new ‘science of, by and for the people’ (Ravetz 2006,277). At the same time, he recognises that ‘[g]iven all thecomplexities and value commitments in the situation, the‘science’ cannot realistically or reasonably be expected to betrivially conclusive for the ‘policy’. [..] Hence the dialogueis not so much one of scientific demonstration, but rather ofnegotiation, where the science is one element among several’(Ravetz 2006, 278). This negotiation is of course the essenceof politics, where strategic behaviour and the use of powerare rife – and this is not always the ‘negotiation in goodfaith’ Ravetz would like to see (2006, 278). While admittingthe limited role of science in decision making he stillconcludes that ‘the theory of Post-Normal Science is nowapproaching obsolescence’ and needs to be replaced by a ‘newbasic insight [..] based on my qualitative version of complexsystems theory’ (Ravetz 2006 275). As in 1993, he identifiesthe reason for societal problems as a scientific misconceptionof what these problems entail and where they originate,assuming that the ‘right’ political choices will be made ifunderrepresented, less powerful groups are invited tocontribute to the search for solutions. However, suchmodification cannot deal with antagonistic strategies amongstpolitical actors as it assumes an environment without power

2 although this is denied by Turnpenny et al. (2009) when they sate that‘[..] democratisation may not have been the original intent of PNS’(Turnpenny et al. 2009, 348)

6

differences, as in Habermas’ communicative ethics andespecially the conditions for ideal speech (Habermas 1981).

However, to declare PNS obsolete because environmentalproblems have not been solved or because scientific practicehas not become as democratic as hoped for seems to us to throwout the baby with the bath water. After all, if ‘the task forthe philosophical critic then was to show that not allproblems with a scientific appearance are capable of solutionin orthodox scientific terms’ (Ravetz 2006 276), this task hasnot gone away today even if ‘uncertainty’ is now an acceptedpart of the scientific and policy discourses – thanks also tothe ideas and efforts of Funtowicz and Ravetz. We believe wecan explain the lack of impact as Ravetz perceived it in 2006from an analysis of the politics of PNS. We thereby concurwith Turnpenny et al. (2009) who, in their analysis of theapplication of PNS, concluded that future researchrequirements on PNS include ‘[..] particularly the need tomore clearly relate theory to different strands of literatureon the evidence–policy-making relationship [..]’ (Turnpenny etal. 2009, 347). We introduce some of this literature here,which enables us to propose a more complete framing of whatPNS sets out to do. However, first we will examine briefly howPNS has fared since its launch in 1993 as this provides inputfor our subsequent analysis.

3 Post-normal science as a new method of doing science?

In their well-known 1993 paper Funtowicz and Ravetz presentpost-normal science (PNS) as a description for a phenomenonthey observe: ‘In response to the challenges of policy issuesof risk and the environment, a new type of science - ‘post-normal’ - is emerging’ (Funtowicz and Ravetz 1993, 739).However, in their 1994 article PNS is a prescription: ‘The newproblems of ecological economics call for a ‘post-normal

7

science’ (Funtowicz and Ravetz 1994a, 197). In Ravetz’ 1999paper PNS is ‘an insight rather than a theory’ which‘justifies itself by its usefulness in helping people tounderstand and manage their problems’ (Ravetz 1999, 642), i.e.PNS is a heuristic. In 2006 Ravetz promotes PNS to a theory(Ravetz 2006, 275) but in the same paper calls it a‘methodology of solving the policy-science problems’ (Ravetz2006, 279). So is PNS a phenomenon, a prescription for a newkind of science, a heuristic, a theory, or something elseagain? We briefly discuss this question by looking at how PNShas been put into practice, i.e. what does it mean to ‘doPNS’. The answer will also start to reveal reasons why PNS isnot as successful as its proponents hoped.

Starting from Funtowicz and Ravetz’ papers, the elements thatseem to be specific to PNS are: 1) to assess whether theproblem is post-normal using the criteria of systemsuncertainties, value contestation, high decision stakes andurgency, 2) to assess and/or improve the quality ofinformation using NUSAP and/or extended peer review. Van deKerkhof and Leroy (2000) suggest that elements specific to PNSare: 1) management of uncertainty, 2) management of inter-and/or transdisciplinarity, 3) management of policy relevance,4) management of quality. In these summaries, we find it hardto detect something that distinguishes PNS from generalcurrent policy analysis theory and practice, except for theuse of NUSAP which is specific to PNS (Hoppe 1999; 2002). As asupplementary approach to finding out what it means to ‘doPNS’ we have therefore used the Web of Knowledge to produce alist of references with the search terms ‘post-normal science’or ‘NUSAP’ in topic or title. The work thus selected wasexplicitly classified as PNS by its authors; this avoids ushaving to produce our own (contestable) criteria for judgingwhether something counts as PNS. Googling the term ‘post-normal science’ did not yield any results beyond academia,except for a recent public conflagration around a Guardian

8

newspaper article by Mike Hulme on climate change science, buthe is an academic and so were his detractors (Hulme 2007)3.

In our rapid assessment of the almost 100 references producedby this Web of Knowledge search, we find on the one handspecific methods and applications for dealing withuncertainty, and on the other more general cases ofdeliberative procedures for dealing with wicked environmentalissues that are described as ‘PNS’. The majority of the papersfalls in this second category. Here, PNS is a ‘sensitizingconcept’ that alerts readers to the need for a PNS approach,either because the problem type is wicked, and/or because theeco-social system is large and complex. Generally in thesepapers there is no indication how exactly PNS can be done,beyond a mention of ‘dialogue’ or ‘stakeholder participation’and/or ‘working at the science-policy interface’. There areother papers with the inverse reasoning: a process of inquiryabout a problem assessed as post-normal is labelled PNS justbecause it involved public participation (e.g. Hisschemölleret al. 2001). Without specifications about the purpose andmodalities of such a set-up it is not clear why this shouldbe named PNS and not a deliberative process like many others,described e.g. by Renn (2008).

We also find examples where PNS is used in a stricter sense:to map, communicate and discuss uncertainties, which may ormay not be related to risks, and/or to improve the quality ofinformation by extended peer review. A few years beforelaunching the concept of PNS, Funtowicz and Ravetz proposedthe NUSAP system to structure the provision of informationabout the different types of uncertainty that affectscientific information (Funtowicz and Ravetz 1990). NUSAPassesses the quality of quantitative information by means of3 This paper was revised when ‘Climategate’ broke. It remains to be seenwhether PNS-type analyses will have any impact on its consequences.

9

five qualifiers: Numeral, Unit, Spread, Assessment, andPedigree (www.nusap.net). A few scientists have worked withthe NUSAP scheme. One of them is Funtowicz himself, who hascontinued elaborating application of PNS through his work forthe European Commission on the precautionary principle(Saltelli and Funtowicz 2005). In the Netherlands NUSAP wasapplied by Van der Sluijs and collaborators (Van der Sluijs etal. 2005). PNS was also used in the Guidance for UncertaintyAssessment and Communication (Jansen et al. 2005) for theDutch National Institute for Public Health and the Environment(RIVM). Its usage, both within RIVM and in communications withits customers, turns out to be limited, probably because theguidelines were not developed and/or discussed with internaland external users (De Vries 2008). This is rather ironicgiven that PNS stresses the involvement of extended peercommunities for producing quality information.4

There may be more profound reasons why policy makers findscientific information about uncertainties unappealing andwhich point to the need to understand the politics of doingPNS: it turns a presentable ‘policy story’ with a beginning,middle and end into a non-story, robbing it of its structurethat was based on facts and effectively disallowing aconclusion (Roe 1994, 5). Funtowicz and Ravetz’s emphasis onuncertainty in the knowledge base could then have the oppositeeffect to what they intended. Merely increasing uncertaintyaround the dominant policy narrative generally triggerspolitical pressures to retain the ‘old’ one (ibid.). Lindblom(1959), Simon (1955) and other early contributors toempirically-based theories of decision making describe howpolicy processes generally deal with complexities,uncertainties, lack of knowledge and disparate values andinterests. Their investigations led to a formulation of atheory of policymaking as intertwined cogitation and4 At present, one of the authors is involved in efforts to remedy thissituation

10

interaction. Lindblom (1965, 1977) concluded that‘[i]nteraction often serves as a social method of analysis[..] an alternative to ratiocination for taking account ofnumerous considerations that ought to be incorporated into anintelligent collective outcome’ (Woodhouse and Collingridge1993, 132). Although Funtowicz and Ravetz favour thisinteraction when they advocate deliberation, they putmodification of scientific methods central in their solutionand thereby implicitly disagree with ‘the lynchpin ofLindblom's thinking [..] that analysis could be, and shouldbe, no more than an adjunct to interaction in political life’(Woodhouse and Collingridge 1993, 132). Rather than to trustthe ‘intelligence of democracy’, however imperfect it may bein contemporary political systems, Funtowicz and Ravetzthereby in effect privilege ‘ratiocination’ overdeliberation.

In conclusion, there are some concrete examples of theapplication of PNS as a method to make uncertainties explicit,although here, too, the relation with policy making is notunproblematic. However, the majority of examples labelled‘PNS’ describes modes of policy analysis that areindistinguishable from other examples of participatory ordeliberative policy-analytic exercises not labelled PNS. Here,PNS is a sensitising concept that alerts to the existence ofcertain problems, rather than a distinct method for dealingwith them. Indeed, more than 20 years after PNS was launched,Turnpenny et al. (2009) conclude that ‘[i]t is ratherdifficult to pin down detailed criteria for assessing thedegree of ‘post-normality’ of activities and practices’(Turnpenny et al. (2009, 350) and proceed to study thepractice of PNS through cases that were not actually explicitattempts to implement a PNS approach but where, in theirassessment, ‘elements of PNS are actually used in practice toaddress wicked issues’ (Turnpenny et al. (2009, 348). Theirfinding that actual PNS applications are rare would explain

11

why in 2006 Ravetz continues to argue that PNS ought to beimplemented, without reviewing actual applications (Ravetz2006).

If PNS was successful as a sensitising concept, this isapparently not sufficient for ‘doing PNS’. To extend the PNStheory, we suggest a possible way forward by examining thepolitics of PNS more closely and from this derive how itsconceptualisation should be revised for it to become moreeffective at achieving its goals.

4 Including politics in a post-normal science framing

We contend that an more comprehensive framing of science inand for policymaking will provide insight into the supposedfailings of PNS to bring about the envisaged policy changes.The fundamental flaws in the implicit PNS conceptualisation ofpolicymaking and the role of science prevent an understandingof the limited usability of a PNS approach to science-for-policy. PNS is in that respect no different from manycontemporary science-for-policy methodologies proposed in andfor environmental problems like integrated assessment,interdisciplinarity, adaptive management, transitions, sociallearning, etc. This means that ‘the art of politics must comeback into the discussion [on environmental problems] if changeis to occur’ (Ingram 2008, 1).

From the above discussion it is clear that PNS frames thesearch for solutions in terms of doing science (after all, itis post-normal science). At the same time, participatory anddeliberative processes and environmental policy choices arewoven into the PNS-as-science framing. Since PNS is steeped inpolitics, to conceptually separate the three activities ofpolicy making, deliberative processes and science (post-normal

12

Governance

New mode of science

uncertainty as incommensurability

mutual understanding

Deliberative democracy

joint problem processing

uncertainty as complexity

uncertainty as ignorance and indeterminacy

extended peer review

PNS

or other) will offer much-needed clarification of theobjectives and possibilities of PNS. This would allow PNS tobecome an approach for understanding and dealing withenvironmental problems.

Figure 1 PNS as science, governance and democracy (adaptedfrom Pellizzoni 2003)

Pellizzoni’s (2003) conceptualisation of participatorytechnology assessment (PTA) as the intersection of practicesof governance, deliberative democracy and (new) science is auseful analogue to clarify PNS (see Figure 1). With theirinterpretation of uncertainty as ignorance and indeterminacyand their demand for extended peer review, Funtowicz andRavetz place PNS firmly in the ‘new science’ compartment. Atthe same time, they present this new science as if it were acomprehensive tool for more democratic and improved governance

13

(centre of Figure 1) because they include deliberation andgovernance issues in the PNS method, based on arguments aboutcomplexity of the knowledge base and fundamental ambiguity andcontestation around values. Unfortunately, while Funtowicz andRavetz were aware that their proposition was politicallycharged (Ravetz 2006), they did not explore the consequencesfor the success of their course of action. Below, weelaborate how our re-conceptualisation explains why PNSencounters resistance, and hence also what it needs to do toadvance its cause.

After briefly discussing well-known pitfalls of theinvolvement of extended peer communities, we focus on thegovernance component of PNS which has received much lessattention. We explore the meta-level of agenda setting wherepriorities are decided, through the governance of problemswhere wicked problems are framed, to the boundary work betweenscience and politics and the balance between power andrationality in problem processing. In doing so, we reflexivelyuse these insights to explain where and why Funtowicz’ andRavetz’ encounter resistance to the application of PNS.

4.1 Post-normal science and deliberative democracy

The proposal to include less powerful actors in policydeliberations ‘exposes and clarifies the political dimensionof specific practice’ by challenging ‘the presumption thatanything ‘scientific’ must be free of uncertainty, independentof values, and the exclusive possession of a technocraticelite’ (Ravetz and Funtowicz 1999, 642). They were not thefirst to conclude this (Hoppe 1999). For example, Weingart haspublished detailed analyses of scientists’ interactions withpolitics since 1970, showing how both worlds get ever moreentangled (e.g. Weingart 1982). In his 1980 essay on thepolitical context of scientific advice, Ezrahi describes how‘scientific knowledge can be drawn upon to serve political

14

ends in contentions about public policy [..] It may very wellmean, particularly when there is disagreement about theobjectives of policy, that science is utilised for itspolitical rather than for its intellectual value’ (Ezrahi1980, 127). Finally, Jasanoff’s seminal 1990 book ‘The fifthbranch: scientists as policy advisors’ deconstructs the samearena of risk analysis that is Funtowicz and Ravetz’ initialconcern.

The political agenda of PNS does not end with this advocacyfor the inclusion of non-experts in deliberations about thequality and policy implications of scientific evidence. Its‘deep political commitments’ (Ravetz 2006 277) are to changesin environmental management, initially formulated as ‘remedythe pathologies of the global industrial system’ (Funtowiczand Ravetz 1993, 739). Ravetz’ formulation of the politicalagenda of PNS has since evolved into the assertion that‘science will need to be done in the cause of justice andsustainability’ (Ravetz 2006, 277). However, just how theinclusion of extended peer communities can or will remedy thepathologies of the global industrial system or deliver justiceand sustainability is not clear from the founders’ texts. Theyrecognise that ‘[p]ublic agreement and participation, derivingessentially from value commitments, will be decisive for theassessment of risks and the setting of policy’ (Funtowicz andRavetz 1993, 751) but they do not elaborate whetherenvironmental protection is part of these commitments, or whatthe effect may be if they are not. After all, ‘environmentalpreferences of the involved actors determine the environmentaloutputs (and outcomes) of decision-making’ (Newig and Fritsch2009, 210). They also seem to imply that big industry andother economic interests do not count as a stakeholders (orthey would overrule environmental concerns); nor do theyexplain how to act if stakeholders have opposite views andpriorities. In fact, by talking of ‘stakeholders’ in generalthey justify the role of business in environmental management,

15

while their explicit purpose is to counter balance the powerof big business.

In a recent workshop, Ravetz recognised these politicalaspects of deliberation. He attributes failures to achievesustainability to the fact that not everybody who should beincluded is invited to contribute: ‘It is hard to avoid theconclusion that we have an ongoing situation of ruthless firmsand toothless watchdogs. [..] The description of the work ofthe Extended Peer Community did mention activist engagement ofvarious sorts, but the need for this was not spelled out’(Ravetz 2008). However, it should be clear that inviting lessprivileged groups is not sufficient to change the course ofpolicy making. In the remainder of this Section we explore theother elements of governance where PNS intervenes withexisting power relations.

4.2 Agenda setting as meta-governance

If, as we assert, PNS has a strong political agenda, theanalysis of its successes and failures should start with theway in which political priorities are decided. Ravetzcorrectly observes that governments have to somehow reconcilecontradictory roles of ‘acting both as promoters of globalbusiness enterprise and also as regulators on behalf of asophisticated and suspicious public’ (Ravetz 2003, 811;original statement of this view in Rip et al. 1995). We agreethat the pathology of the present global industrial system isits relentless drive for profits through market-typeefficiency which disregards external costs: mass migration,damage to health, damage to nature, and many more. It isexactly because of the incompatibility of the political-economic and biophysical paradigms that PNS-type extended peerreview could come in as a way for the governance structure ofhuman societies to successfully adapt to changingcircumstances. However, as we have already discussed, it is no

16

foregone conclusion that deliberative processes will indeedprioritise sustainability over economic interests. Second, thepolitical choice for deliberation is not shared by relevantothers. Many politicians, political scientists and citizensbelieve that representative democracy is superior todeliberative democracy exactly because it is better adapted tocomplex, large-scale socio-political systems. Also, not onlyneo-conservatives and neo-liberals believe the market systemto be superior as a societal coordination mechanism exactlybecause it handles the complexity of (economic) transactionsin coordination between billions of people all over the globebetter than systems of central coordination (Lindblom 2001).It is hardly surprising that these scientists who prescribe amove towards deliberative democratic procedures in order toadapt better to the complexities of the current socio-economic-politico-ecological system run into resistance fromvested interests and well-engrained institutions as they implythat political, economic and scientific elites ought to cedeparts of their power and control to others. This is a realitythey need to take into account, even if those same elitesapparently subscribe to the arguments for change.

Ironically, while exploiting the idea of bio-physicalcomplexity Funtowicz and Ravetz’ argument for extended peerreview therefore disregards the complexity of moderngovernance. Without explicit attention to institutionalalignment such ideas will never be successfully ‘exported’ togovernance structures outside science. Limiting ourselves tothe institutions of political democracy5, processes ofdeliberative democracy have to be aligned with representativedemocracy and neo-corporatist policymaking in many policysubsystems. Goodin and Dryzek (2006) call this the problem of

5 For a full picture one would have to include complex and volatile multi-institutional alignments with markets, firms, civil society organizationsand associations, and individuals as consumers, citizens, and members ofhouseholds.

17

‘macro-political up-take of mini-publics’ while Edelenbos etal. (2008) frame this as ‘democratic anchorage of interactivegovernance’. While Funtowicz and Ravetz aim to integratedifferent values and represent stakes in extended peercommunities, this is not sufficient to ensure that subsequentpolicy decisions take account of the results. On the contrary,in the view of many politicians including non-experts inscientific peer review means less rather than more authorityand credibility. In that sense, PNS may have been counter-productive in terms of its own expressed intentions.6 Moreover,many experiments with deliberation are actually commissionedby decision makers who owe this position and role to thepower-mechanisms in representative democracy (e.g. Loeber2004). ‘Doing it deliberatively’ is just one option out ofmany for these authorities in getting decisions and policy outof the political ‘machine’. By keeping open the option forthemselves to not even respond to recommendations, they givethe impression of not taking seriously procedures they havethemselves set in motion (Joly and Marris 2003). Small wonderthat many advocates of deliberative procedures, Funtowicz andRavetz included, get disappointed about the lack of changesbrought about by such experiments.

What is actually needed at this level of agenda-setting is asustainability meta-governance that covers the full complexityof institutional alignments. So far we can only hope that thiswill happen, as economic concerns remain paramount. AsBaumgartner and Jones (1993) say in their theory of long-termpolicy dynamics: under conditions of rapid change and entranceof new policy players old policy subsystems crumble and new

6 The scientists who vehemently protested at Mike Hulme’s assertion thatclimate change science is post-normal have understood that this claimdiminishes their authority. See the article (Hulme 2007) and thesubsequent exchange:http://www.guardian.co.uk/society/2007/mar/14/scienceofclimatechange.climatechange See also Kysar (2008).

18

policy networks may emerge. But apart from hoping for this tohappen spontaneously, perhaps some deliberate politicalsteering is possible in the form of ‘institutionalentrepreneurship’ (Hoppe 2010). The concept was coined byDiMaggio (1988) who held that ‘new institutions arise whenorganized actors with sufficient resources see an opportunity[..] to realize interests that they value highly’ (DiMaggio1988, 14). Policy entrepreneurs are nudging the existingpolicy network in a more desired direction by reframing thepolicy problem and rethinking the conditions for this newframe to become legitimate. With their propositions to reframeenvironmental problems and their emphasis on urgency anddemocracy Funtowicz and Ravetz are trying to do this. Thetrouble with PNS is that science cannot jump over its ownshadow: it is part and parcel of the complex governancestructure which it tries to influence, while the translationwork of institutional entrepreneurs depends on their locus atthe margins or ‘above’ policy networks. Being familiar withmore than one policy framing and policy-political logicenables them to think more innovatively than the network‘insiders’. In addition, they can draw on financial andcommunicative resources outside the normal network structuresto bolster their efforts.

However, policy entrepreneurship is by no means a sufficientcondition for bringing about the desired changes. The systemis simply too complex and self-controlled to be amenable tometa-governance by powerful leaders. In societies like ours,with complex, overlapping, partially contradictory governancesubsystems, the only way to ‘steer’ is by using the forces anddrivers already present: to ‘interpolate’ small doses ofchange in such ways that the balance of forces is changed inthe desired direction (Dunsire 1986, Hood 1986). This policy-cum-institutional entrepreneurship is an essential ingredientof a political system in the responsive and creativegovernance of problems. At this level of meta-governance, the

19

value of PNS as sensitizing concept is in sparking off bitsand pieces of institutional and policy entrepreneurship whichmay, ultimately, build up to a new governance structure fullyadapted to the threatening alterations in society’sbiophysical environment. However, this is insufficient initself.

4.3 The governance of problems and problem types

In addition to their activities as policy entrepreneurs,Funtowicz and Ravetz are also implicitly trying to exertinfluence at the level of the politics of problem framing.Here, problems are deliberately and strategically framed inorder to mobilize political and social support(Schattschneider 1988). We will show that conceptually theytry to redefine the problem type and thereby also the way itis dealt with. In doing this they challenge well-establishedpolicy networks and practices, thus encountering resistance atthis level. too. Our analysis now supposes that sustainabilityhas indeed made it onto the political agenda and has beenidentified as a policy problem requiring some form ofcollective response, as is indeed the case in westernsocieties. The question then becomes: how to arrive at thiscollective response, i.e. how to go about developing policies?Generally speaking, this depends on how powerful decisionmakers structure the problematic situation. Using the socialstudy of organizations (Thompson and Tuden 1957), science andtechnology (Ezrahi 1980), and public policy (Van de Graaf andHoppe 1989), Hisschemöller and Hoppe (1996) have captureddifferent ways of structuring problems in a well-knowntypology (Figure 2). It rests on a definition of a problem asa deviation between an ‘is’ and an ‘ought’. The ‘is’ is theavailable relevant knowledge for understanding the problem, inwhich there can be more or less certainty. The ‘ought’ isrepresented in the norms, values, ideals and interests atstake in defining the problem, in which there can be more or

20

less ambivalence or ambiguity. These two dimensions yield fourproblem types.

Figure 2 Typology of problem structures (Hisschemöller and

Hoppe 1996)

Participants in policymaking generally try to steer debatesinto the quadrant where they can win the dispute, or whereopponents are not likely to win (Schattschneider 1988). Inthis problem structuring, implicit and explicit choices aremade about what to include and what to ignore, i.e. theproblem is limited in scope by selecting a particular frame.This means that problem structuring moves a moderately orunstructured problem into one of the other quadrants, afterwhich an appropriate, sometimes scientifically endorsedstrategy will enable dealing with the problem, at leasttemporarily. This enables actions to be taken towardssolution, although frequently it cannot be expected that theseare more than temporary settlements which bring onlyincremental improvements (Lindblom 1959, 1965; Hayes, 2001).

21

Each problem type has a distinct politics of policymaking andconcomitant problem processing strategies (Figure 3).Structured problems are solved by a debate on technicalitiesof fine-tuning the means to achieve objectives; moderatelystructured problems with shared goals lead to debate on the(evaluation of) a set of means. Structured problems areattractive to politicians and policymakers because they have atrack record of ‘do-ability’ (Rittel and Webber 1973): afeasible, well-understood method exists that satisfies allcriteria for an adequate, successful solution. This is exactlythe function of ‘normal’ science as problem solver. Even ifthey realize that most politically salient problems are notquite fully structured, policymakers want to move problemsinto the quadrant of structured problems rather sooner thanlater so they are seen to have solved the problem (March andSimon 1993, Schön 1983).

Figure 3 Problem typology and politics of policymaking (Hoppe2010)

22

By arguing that certain problems are ‘post-normal’ Funtowiczand Ravetz engage in the governance of these problems. Theyargue against ‘normal’ do-ability when they assert that ‘factsare uncertain, values are in dispute, stakes are high anddecisions urgent’ (Funtowicz and Ravetz 1993, 744).Conceptually they are thus trying to move environmental orsustainability problems that were considered (moderately)structured clean-up or externalities problems by policymakersand scientists alike into another quadrant. Any such effort atre-classification challenges accepted problem solvingstrategies and thereby the division of labour amongparticipants in the policy network that has emerged to performthis. In the iterative policy process of problem framing andsolving policy players attempt to gain authoritativeacceptance for their claims through persuasion, bargaining andother processes of partisan mutual adjustment (Figure 4). Theserial character of rounds of policymaking can lead to theinstitutionalization of certain problems as ‘permanently’structured, where one problem framing becomes the hegemonicproblem definition.

Policy implementers puzzle about alternative solutions withinthe bounds of the set of constraints implied by the dominantproblem definition, and professional or academic advisers makediscoveries about the pros and cons of the constraints. Theresulting learning processes, and/or external politicalevents, may lead to novel critical moments, which may lift(some) older constraints, or introduce new ones. Thehistorical sequence of such processes creates pathdependencies in problem definitions and constraint sequencingas well as a policy network that dealt with them. Such pathdependencies either facilitate or work against the adoption ofnovel problem/solution couplings. This explains part of theresistance that Funtowicz and Ravetz have experienced intrying to bring about changes in the dominant problemdefinition, while this framing is critically important as

23

‘[a]mong the critical shortcomings in contemporary[environmental] politics are the failures to frame issues inways to attract public interest, to engage a[n environmental]ethic and to address deep seated inequities, to recruit and toinspire leaders’ (Ingram 2008,1).

In addition, we question whether Funtowicz and Ravetz placeenvironmental problems in the correct problem category whenthey propose their framing. Although they mention that ‘valuesare in dispute’, they at the same time seem to assume thatthere is societal agreement on the need to deal with theseproblems urgently. The problem type they thereby assume isreflected in the category of moderately structured problemsfor which there is agreement on goals (MSP(g) in Figure 3).However, we believe that post-normal problems should beconsidered unstructured (UP in Figure 3). There is noagreement in society on a priority for sustainability overe.g. economic development, and there is certainly no agreementover what sustainability entails. Funtowicz and Ravetz’implicit classification of post-normal problems as MSP(g) canexplain further the disappointing impact of PNS. Whereas UPproblems demand efforts like an agonistic politics of agenda-changing, where possible combined with learning throughdeliberation and participation (Figure 3), MSP(g) problems canoften be solved by relatively ‘normal’ problem-drivensearching and negotiation. With their preference for‘ratiocination’ they employ the wrong-problem processingstrategy: one that fits MSP(g) and not the UP that post-normalproblems represent.

24

Figure 4 How dominant problem definitions come about andchange

(adapted from Vergragt 1988, 503)

4.4 Policy politics and the role of scientific information

It is clear that problem structuring is a politico-cognitivenecessity in dealing with unstructured and moderatelystructured problems. Problem structuring may be dominated by apuzzling or powering style. Dunn (2007, 6) defines thepuzzling style as the self-conscious search, analysis andevaluation of competing problem representations and problemframings, with a view to their possible integration anddefinition. In a more political twist, problem structuringcould be defined as the political activity to produceinformation on divergent views of what the problem is about,with a view to produce a synthetic, or at least politicallyplausible, problem definition (Hisschemöller and Hoppe 1996).In practice, problem structuring is, at best, an oscillationbetween ratiocination and rationalization. Much more likely,it is a complex two-level real-time knowledge/power game; likethe next move on the chess board being determined by the

25

outcome of a boxing match, and vice versa. This dual nature ofproblem structuring as cogitation and political interaction,as powering and puzzling, is inevitable because determiningwhether or not to open up or close down a debate, and whetheror not to allow the closure of particular constraints, alwaysentails political struggle. This is especially true forunstructured problems because there are no logical orinherently ‘rational’ criteria for beginning or endinginfinite disputes.

Problem structuring as puzzling looks identical to thedeliberative processes proposed by Funtowicz and Ravetz. Intheir original conceptualization they believed that externalpeer review could be sufficiently insulated against theinfluence of ‘politicking’. Yet this flies in the face ofexperience. Those in power define what counts as ‘rationality’and knowledge. The relationship between power and rationalityin problem framing is asymmetrical: ‘[..] power has arationality that rationality does not know. Rationality, onthe other hand, does not have a power that that power does notknow’ (Flyvbjerg 1998, 234). In the formulation of Ezrahi(1980) there is a ‘preponderance of political considerationsin the formulation and execution of public policies. Theprocess of democratic politics consists of endless conflicts,negotiations and temporary compromises; legislatures andpublic bureaucracies which [..] are not under compellingpressure to integrate scientific knowledge and technicalstandards into their activities. [..] It is our task tounderstand how scientific knowledge does become woven into thepolitical and bureaucratic formulation and execution ofpolicies’(Ezrahi 1980 111). Following Ezrahi’s suggestion totrace the ways in which scientific information is input intothe process under different conditions of policy politics.Based on empirical studies of dozens of Dutch cases,Hisschemöller et al. (2001) proposed the following typology

26

for the role of science as dependent on problem types (Figure5).7

Figure 5 Problem typology and the role of science(Hisschemöller et al. 2001, 464)

4.5 Boundary work on the science-politics interface

If the Hisschemöller et al. (2001) typology is generallyvalid, the implication of different problem typologies is that‘the [authoritative] problem structure limits theopportunities for scientists to shape their own roles inenvironmental policymaking, as they are bound to play aspecific role either as problem solver, as advocate, asmediator, or as problem finder. Therefore, science use dependson the structure of the problem as constructed by dominant7 Note that Hisschemöller et al. (2001) use low-profile descriptive labelsto describe the roles of scientists, and avoid high-level normative labelslike Pielke’s (2007) ‘honest broker’ or ‘stealth advocate’, or Jasanoff’s(2003) revival of Mertonian norms as ‘technologies of humility’.

27

policy actors at a given moment, and the informal rules of thegame that drive the policy process’ (Hisschemöller et al.2001, 465). From descriptive studies of boundary work andboundary organizations we know that the relation betweenexperts and policy makers is a complex and contested divisionof labour. This division of labour consists of a boundary thatdemarcates who can and cannot be considered an expert invarious degrees, and articulates the coordination betweenactors who have come to be considered ‘experts’ and ‘policymakers’. Such boundaries are the outcome of – and form theresources for – continuing boundary work, the furtherarticulation, reproduction, or modification of this divisionof labour (Halffman and Hoppe 2006, 135). While theseconstraints clearly exist, this does not mean that scientistsdo not try to escape from them, as Funtowicz and Ravetz show.PNS is relatively successful in a role as problem solver (SPsquare) and problem analyst (MSP-g) through its method forassessing uncertainties such as NUSAP, or as problemrecogniser (UP square), with PNS as sensitising concept. Theseare then apparently roles that are accepted in the currentboundary arrangements around environmental problems. However,when they try to take a role as advocate (MSP-g square) ormediator (MSP-m square), Funtowicz and Ravetz encounteropposition: these roles are apparently not yet part of theaccepted arrangements, and to change this requires politicalstruggle and gradual institutionalisation.

An understanding of the political agenda of PNS from aboundary work perspective is thereby the last element of ourexplanation of the resistance encountered by Funtowicz andRavetz and other PNS enthusiasts. In trying to escape from thelimits to their role imposed by the dominant problem framingand its resultant prescriptions, they unwittingly engage inboundary work between science and policy because as scientiststhey ‘transgress’ onto the political side of the boundary.Hoppe (2005) describes eight theoretical boundary arrangements

28

which he classified on the basis of two dimensions: whetherthe primacy lies with science or politics, and whetherscience and politics are seen to converge or diverge in theiressential social functions. He then studied how Dutch policyworkers themselves conceive of the division of labour betweenscience and politics (Hoppe 2009). He found considerableoverlap between the theoretical and the actual discourses. Insustainability-related policy domains (such as agriculture,environment, water management) policy workers could nearly allbe classified as ‘post-normalists’ (which is more general thanPNS as a specific concept). These policy workers believe inthe need for and possibility of convergence between politicsand science in order to serve society, i.e. in solvingenvironmental problems. They expressly acknowledge the primacyof politics, while simultaneously claiming that science is notjust serviceable but has an autonomous responsibility. Post-normalists’ framing of their policy work is therefore full ofcontradictions: on the one hand, they distrust democracy toset the ‘right’ priorities (science is autonomous) and on theother hand they insist on extensive deliberation to bringabout desired changes in environmental policy because theyrealise that they depend on politics to make the rightdecisions (primacy of politics). These are also thecontradictions apparent in Funtowicz and Ravetz’ presentationof PNS.

To find many post-normalists amongst experts insustainability-related policy areas is no Dutch coincidence.In the US and Canada too, it has been found that scientistsworking in sustainability-related policy domains hold viewsthat are best described as ‘post-normal’ (Steel et al. 2004).This general finding results from a conjunction of threecauses: complexity in the knowledge base, instability ofpolicy networks, and lack of political clout. First, thephysical systems that sustainability-related boundary workerswork with are now widely recognised to be characterised by

29

complexity, which results in uncertainty about andunpredictability of information. The complexity also generatesa need to work with other disciplines. Both factors underminethe boundary workers’ substantive authority. Second, post-normalists operate in relatively new policy networks whererelationships are anything but settled, not within thetraditional government structures, nor with an expandingnumber of civil society organizations and stakeholders. Third,sustainability-related issues tend to be low on the list ofpolitical priorities, if not in rhetoric then in practice,with economical interests usually dominating. These threefactors create an uncertain position for post-normal policyworkers. They try to countervail this uncertainty by formingalliances, hence the emphasis on trust in science-politicsdialogue and deliberation, and by re-asserting that theiractivities are solid science, hence the frequent reference towell-established rules and methods (Hoppe 2009, 16). Funtowiczand Ravetz’s insistence that ‘post-normal science is indeed atype of science, and not merely politics or publicparticipation’ (Funtowicz and Ravetz 1993, 751) is an exampleof the latter.

5 What next?

We have shown in this paper that Funtowicz and Ravetz wereaware that their promotion of PNS had a political agenda, butjust how much they did not recognise. Conversely, the PNSapproach is not politicised enough to do what they really wantto do: to further environmental goals, especiallysustainability. So to answer the question asked in the titleof this paper, ‘If post-normal science is the solution, whatis the problem?’: as it stands PNS can be a solution for flawsin a scientific method, but not for the dual problem ofdemocratisation of, and priority setting for, environmentalpolicy making. Post-normal experts represent disciplines whereuntil recently enlightenment and technocracy models enjoyedmuch popularity among academic scientists, and bureaucracy

30

models governed their practices as boundary workers in theservice of politics and policy. The political salience andsocietal relevance of sustainability issues is now pressingthem to find other, workable models for boundary arrangements.Their advocacy of, and discourse coalition building for,policy on the basis of ecological insights may even expressitself as technocracy (Hoppe 2009). In other words: althoughpost-normal policy workers articulate the need for convergencewith politics, they try to make sure that this happens ontheir own terms, not according to democratically set politicalpriorities. Ultimately this means that they enter thepolitical arena as ‘stealth advocates’ (in Pielke’s (2007)terms), but through the back door of science, to findthemselves facing political contestation they did notanticipate. For it should be clear that it is a misconceptionto consider the role of science in policymaking well-defined,either in the normal science or in the post-normal variant. Itis this understanding of the variability in science’s role inpolitics and public policy that is lacking in PNS: Funtowiczand Ravetz are stuck in ‘disillusionment with the power ofscientific knowledge to transform and rationalise thedecisions and actions of governments’ (Ezrahi 1980, 111). Theypropose to remedy certain ‘faults’ of scientific practicewhereas in fact it is the politics and policy of agendasetting, problem framing and solving and in boundaryarrangements where resistances to PNS originate.

Overall, PNS has been much more successful as sensitizingconcept than as fully fledged theory or as heuristics for thegovernance of concrete problems. This is understandable asthere is no scientific way to ‘solve’ unstructured societalproblems. This is a vital observation in view of the many newtools and methods advanced in environmental sciences thatclaim to do just that. However, it is possible to make PNS andother interactive and deliberative methods more politicallysophisticated, and we actually see these changes starting to

31

happen (e.g. Pellizzoni 2003, Turnpenny et al. 2009). Changesshould focus on all aspects of making sustainabilitygovernable: 1) the conduct of deliberative policy analyses andits links to the institutions of representative democracy, 2)the boundary work at the science-policy interface, and 3) theadvocacy coalition and public persuasion work needed foragenda-setting and societal change. Each of these merits aseparate paper but we have space only for a few generalcomments here.

The literature on public participation and deliberation isextensive and there is a well-stocked and ever-increasingtoolkit of methods available. The recommendations to emergefrom evaluations of these processes invariably concern threeinterrelated issues: fostering conditions for real learning,maintaining some sort of power balance between participantswith sometimes very unequal resources in the real world, andmacro-political up-take in the real world of powerdifferentials in representative democracy and capitalism(Loeber 2004, Edelenbos et al. 2008). The substantialliterature on boundary organizations and boundary workprovides more strategic lessons for PNS proponents: gettingyour view to count is all about balancing processes ofdemarcation (for science’s credibility and authority) andcoordination (with a view to relevance and salience) (Halffman2003, Hoppe 2005). These lessons can be briefly summarized asinstitutionalized double participation and accountabilityacross the science-policy divide, and co-production of asocial and cognitive order e.g. through the use of boundaryobjects in continuous communication and interaction betweenpolicy workers and scientific experts (Star and Griesemer1989, Guston, 2001, Miller 2001, Hoppe & Huijs 2003, Nutley etal. 2003, Wesselink 2009).

32

Both participatory-deliberative policy analysis and boundarywork are embedded in political environments where substantiveknowledge about policy problems competes with politicalconcerns: how harmful is the information for someone’s powerposition, how many people believe the problem to be important,how much political support is mobilized through tackling theproblem, what are its financial and other types of social andpolitical costs relative to competing problems (Webber 1992,Nutley et al. 2003). Whether and how it is possible to achieveany of the recommendations in a given situation ultimatelyremains a matter of advocacy, convincing, context-sensitivepolitical judgment and political struggle. Even if it istherefore impossible to present universally applicablerecommendations, it is important that the political nature ofthese issues is clear. There is no use for lists of wishful‘ought to’ recommendations if reality does not comply all byitself: creating space for deliberative experiments,persuading policy makers to listen to scientific findings, andtransitions to a fairer and more sustainable world all requirepolitical commitment and action. Claiming to have the truerfacts and right numbers is only a secondary, though ofteninstrumentally useful matter.

6 Acknowledgements

We much appreciate comments from two anonymous reviewers whohelped us to present our arguments more clearly. This researchwas financially supported by the European Union (EuropeanCommission, Marie Curie RTN GoverNat, contract 0035536,www.governat.eu).

7 References

Baumgartner F R, and Jones B D (1993) Agendas and Instability inAmerican Politics Chicago University Press, Chicago and London

33

Cash D, Clark W, Alcock F, Dickson N, Eckley N and Jäger J(2002) Salience, Credibility, Legitimacy and Boundaries:Linking Research, Assessment and Decision Making. John F. KennedySchool of Government Faculty Research Working Paper RWP02-046. John F.Kennedy School of Government, Harvard University.

De Vries A (2008) Towards do-ability PhD thesis University ofTwente.

DiMaggio P (1988) Interest and agency in institutional theoryZucker L G (ed.) Institutional Patterns and Organizations: Culture andEnvironment Ballinger, Cambridge MA 3-22

Dunn W N (2007) Public policy analysis. An introduction 4th ed. PrenticeHall, Englewood Cliffs

Dunsire A (1986) A cybernetic view of guidance, control andevaluation in the public sector Kaufman F X et al (eds.)Guidance, Control, and Evaluation in the Public Sector Walter de Gruyter,Berlin and New York 327-346

Edelenbos J, van Schie N and Gerrits L (2008) Democraticanchorage of interactive governance: developing institutionalinterfaces in water governance. Proceedings 2008 conference of thePolitical Science Association http://www.psa.ac.uk/cps/cps.asp

Ezrahi Y (1980) Utopian and Pragmatic Rationalism: ThePolitical Context of Scientific Advice Minerva 18 (1) 111-131

34

Funtowicz S O and Ravetz J (1990) Uncertainty and quality in science forpolicy Kluwer, Dordrecht

Funtowicz S O and Ravetz J R (1993) Science for the Post-Normal Age Futures 25(7) 735-755

Funtowicz S O and Ravetz J R (1994a) Uncertainty, complexityand Post-Normal Science. Environmental Toxicology and Chemistry 13(12) 1881– 885.

Funtowicz S O and Ravetz (1994b) The worth of a songbird:ecological economics as a post-normal science Ecological Economics10 197-207

Funtowicz S O and Ravetz (1994c) Emergent complex systemsFutures 1994 26(6) 568-582

Gallopin, G.C., Funtowitz, S., O’Connor, M. and Ravetz, J

(2001) Science for the 21st century: from social contract toscientific core. International Social Science Journal 53:219-229.

Goodin R E and Dryzek J S (2006) Deliberative Impacts: TheMacro-Political Uptake of Mini-Publics Politics and Society 34 (2)219-244

Guston D (2001) Boundary Organizations in Environmental Policyand Science: An Introduction Science, Technology and Human Values26(4) 399-408

35

Habermas J (1981) Theorie des kommunikativen Handelns Frankfurt amMain, Suhrkamp

Halffman, W. (2003) Boundaries of Regulatory Science University ofAmsterdam, Amsterdam (dissertation)

Halffman W and Hoppe R (2006) Science/policy boundaries: achanging division of labour in Dutch expert policy advice. InMaasen S and Weingart P (eds) Democratization of Expertise? ExploringNovel Forms of Scientific Advice in Political Decision-Making (Sociology of theSciences Yearbook) Springer 135-152

Hayes M T (2001) The limits of policy change. Incrementalism, worldview, andthe rule of law Georgetown University Press, Washington D.C.

Hisschemöller M and Hoppe R (1996) Coping with IntractableControversies: The Case for Problem Structuring in PolicyDesign and Analysis Knowledge and Policy: The International Journal ofKnowledge Transfer and Utilization 8(4) 40-60.

Hisschemöller M, Hoppe R, Groenewegen P and Midden C J H(2001) Knowledge use and political choice: a problem-structuring perspective on real life experiments in extendedpeer review. Hisschemöller M, Hoppe R, Dunn W N and Ravetz J R(eds) (2001) Knowledge, power, and participation in environmental policyanalysis Policy studies review annual 12. Transaction, NewBrunswick

36

Hood C (1986) Concepts of control over bureaucracy: ‘comptrol’and ‘interpolable balace’ Kaufman F X et al (eds.) Guidance,Control, and Evaluation in the Public Sector Walter de Gruyter, Berlin andNew York 765-783

Hoppe R (1999) Policy analysis, science and politics: from‘speaking truth to power’ to ‘making sense together’ Science andPublic Policy 26(3) 201-210

Hoppe R (2005) Rethinking the science-policy nexus: fromknowledge utilization and science technology studies to typesof boundary arrangements Poiesis and Praxis 3(3) 199-215

Hoppe R (2009) Scientific advice and public policy: expertadvisers’ and policymakers’ discourses on boundary work. Poiesisand Praxis 6 (3-4) 235–263

Hoppe R (2010) The Governance of Problems. Puzzling, Powering, ParticipationBristol: Policy Press

Hoppe R and Huijs S (2003) Werk op de grens tussen wetenschap en beleid:paradoxen en dilemma's. Den Haag, Raad voor Ruimtelijk, Milieu- enNatuuronderzoek (RMNO): 56.

Hulme M (2007) The appliance of science The Guardian 14 March2007

http://www.guardian.co.uk/society/2007/mar/14/scienceofclimatechange.climatechange

37

Ingram H (2008) Beyond Universal Remedies for Good WaterGovernance: A Political and Contextual Approach. Paperpresented at the Rosenberg Forum for Water Policy, Zaragoza,Spain, June 25-26; http://rosenberg.ucanr.org/documents/V%20Ingram.pdf.

Jansen P C (2007) Beoordeling van de kwaliteit van hoogwaterparametersberekend voor de Nederlandse Maas MSc thesis, department of WaterEngineering and Management, University of Twentehttp://purl.org/utwente/e704

Janssen P H M, Petersen A C, Van der Sluijs J P, Risbey J Sand Ravetz J R (2005) A guidance for assessing and communicatinguncertainties. Water Science and Technology 52(6) 125-131

Jasanoff S (1990) The fifth branch: science advisers as policy makersHarvard University Press Cambridge MA

Jasanoff S (2003) Technologies of humility: citizenparticipation in governing science Minerva 41(3) 223–244

Joly P-B and Marris C (2003) La participation contra lamobilization? Une analyse comparée du débat sur les OGM enFrance et au Royaume Unie Revue Internationale de Politique Comparée, 10(2) 195-206

Loeber A (2004) Practical wisdom in the risk society. Methods and practice ofinterpretive analysis on questions of sustainable development PhD thesis,University of Amsterdam

38

Lindblom C E (1959) The Science of 'Muddling Through' PublicAdministration Review 19 79-88.

Lindblom C E (1965) The Intelligence of Democracy New York: The FreePress.

Lindblom C E (1977) Politics and Markets: The World's Political- EconomicSystems New York: Basic Books

Lindblom C E (2001) The Market System: What it is, How it Works and What toMake of it Yale University Press

Lindblom C E and Woodhouse E J (1993) The Policy-Making ProcessPrentice Hall, Englewood Cliffs NJ (3rd edition)

March J G and Simon H A (1993) Organizations Blackwell, Oxford(1st ed. 1958)

Miller C (2001) Hybrid Management: Boundary Organizations,Science Policy, and Environmental Governance in the ClimateRegime Science, Technology and Human Values 26(4) 478-500

Newig J and Fritsch O (2009) Environmental Governance:Participatory, Multi-Level - and Effective? Environmental Policy andGovernance 19(3) 197-214

39

Nutley S, Walter I and Davies H T O (2003) From Knowing toDoing. A Framework for Understanding the Evidence-Into-Practice Agenda Evaluation, 2003, 9, 2, pp. 125-148

Kysar D (2008) Democratizing knowledge generation: Anintroduction to key issues Conference on Environmental Governance andDemocracy, May 10-11, 2008, Yale University, New Havenhttp://www.yale.edu/envirocenter/envdem/agenda.htm accesses 5February 2010

Pellizzoni L (2003) Uncertainty and Participatory DemocracyEnvironmental Values 12(2) 195–224

Pielke R A (2007) The honest broker : making sense of science in policy andpolitics. Cambridge University Press

Price D K (1967) The Scientific Estate Harvard University Press,Cambridge MA

Ravetz J R (2003) A paradoxical future for safety in theglobal knowledge economy Futures 35 (8) 811–826

Ravetz J R (2004) A Critical Scrutiny of Post-Normal Science.International conference Uncertainty and Precaution inEnvironmental Management, Copenhagen, June 7-9 2004http://upem.er.dtu.dk/programme.htm (abstract)

Ravetz J R (2006) Post-Normal Science and the complexity oftransitions towards sustainability Ecological complexity 3 275-284

40

Ravetz J R (2008) Post-normal science in action. ‘Mixingthings up: science, politics and lay knowledge’, Centre forthe Study of Environmental Change and Centre for the Economicand Social Aspects of Genomics, Lancaster University, 19 March2008 (abstract)

Ravetz J R and Funtowicz (1999) Post-Normal Science - aninsight now maturing Futures 31 (7) 641-646

Renn O (2008) Risk Governance: Coping with Uncertainty in a Complex WorldThe Earthscan Risk in Society Series, Earthscan, London

Rip A, Schot J W and Misa T J (1995) Managing Technology in Society.The Approach of Constructive Technology Assessment Pinter Publishers,Londen/New York

Rittel H W J and Webber M D (1973) Dilemmas in a GeneralTheory of Planning Policy Sciences 4 155-169

Roe E (1994) Narrative Policy Analysis. Theory and Practice. Durham andLondon: Duke University Press

Saltelli A and Funtowicz S (2005) The Precautionary Principle:Implications for Risk Management Strategies Human and EcologicalRisk Assessment 11(1) 69–83

Schattschneider E E (1988) The Semi-Sovereign People. A Realist’s View ofDemocracy in America Wadsworth Thomson Learning (1st edition 1960)

41

Schön D A (1983) The reflective practitioner: how professionals think in actionNew York Basic Books

Simon H A (1955) A Behavioral Model of Rational Choice QuarterlyJournal of Economics 69 99-118.

Shackley S, Wynne B and Waterton C (1996) Imagine complexity:The past, present and future potential of complex thinkingFutures 28(3) 201-225

Star L S and Griesemer J R (1989) Institutional ecology,‘transitions’ and boundary objects: amateurs and professionalsin Berkeley’s museum of vertebrate zoology Social Studies of Science19 387–420

Steel B, List P, Lach D, Shindler B (2004) The role ofscientists in the environmental policy process: a case studyfrom the American west Environmental Science and Policy 7 1–13

Thompson J D and Tuden A (1959) Strategies, Structures, andProcesses of Organisational Decision. J.D. Thompson, P. B.Hammond, R.W. Hawkes, B.H. Junker, A. Tuden (eds.) Comparativestudies in administration University of Pittsburgh Press Series incomparative administration nr. 1

Turnpenny J, Lorenzoni I and Jones M (2009) Noisy anddefinitely not normal: responding to wicked issues in theenvironment, energy and health Environmental Science and Policy 12(3)347-358

42

Van de Graaf H and Hoppe R (1989) Beleid en politiek: een inleiding tot debeleidswetenschap en de beleidskunde Coutinho, Muiderberg

Van de Kerkhof M and Leroy P (2000) Recent environmentalresearch in the Netherlands: towards post-normal science?Futures 32(9-10) 899–911

Van der Sluijs J P, Craye M, Funtowicz S, Kloprogge P, RavetzJ and Risbey J (2005) Experiences with the NUSAP system formultidimensional uncertainty assessment Water Science and Technology52(6)133–144

Vergragt P J (1988) The social shaping of technologicalinnovations Social Studies of Sciences 18 (3) 483-513

Webber D J (1992) The Distribution and Use of Knowledge in thePolicy Process, edited by Dunn W N and Kelly R M (eds) Advancesin Policy Studies since 1950. New Brunswick and London: TransactionPublishers 383-418

Weingart P (1982) The scientific power elite - a chimera; thede-institutionalization and politicization of science. NElias, H Martins and R Whitley (eds) Scientific Establishments andHierarchies. Sociology of the Sciences Yearbook vol. 6 Reidel,Dordrecht 71–87.

Wesselink A J (2009) The emergence of interdisciplinaryknowledge in problem focused research Area 41(4) 404-413

43

Woodhouse E J and Collingridge D (1993) Incrementalism,intelligent trial-and-error, and political decision theory.Harry Redner (ed.) An Heretical Heir of the Enlightenment. Science, Politics andPolicy in the Work of Charles E. Lindblom. Boulder, CO: Westview. 131-154

44