UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Imbued meaning : science-policy interactions in the intergovernmental panel on climate change Shaw, Alison 2005

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_2005-105674.pdf [ 16.11MB ]
Metadata
JSON: 831-1.0092354.json
JSON-LD: 831-1.0092354-ld.json
RDF/XML (Pretty): 831-1.0092354-rdf.xml
RDF/JSON: 831-1.0092354-rdf.json
Turtle: 831-1.0092354-turtle.txt
N-Triples: 831-1.0092354-rdf-ntriples.txt
Original Record: 831-1.0092354-source.json
Full Text
831-1.0092354-fulltext.txt
Citation
831-1.0092354.ris

Full Text

IMBUED MEANING: SCIENCE-POLICY INTERACTIONS IN THE INTERGOVERNMENTAL PANEL ON CLIMATE CHANGE by ALISON SHAW Honours B.A., The University of British Columbia^ 2000  A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in FACULTY OF GRADUATE STUDIES (Resource Management Environmental Studies)  THE UNIVERSITY OF BRITISH COLUMBIA May 2005  © Alison Shaw, 2005  Abstract The proliferation of scientific information in the international policy sphere has increased with the proliferation of global environmental problems. The conventional transfer of scientific information becomes increasingly complex in the international sphere where the implications of global environmental problems are severe and where divergent values around the type of information considered sufficient and adequate for policy action lead to differentiated governmental responses. Constructivist science-policy scholarship has challenged the unidirectional transfer of science into policy suggesting that the sociopolitical sphere plays a significant role in determining the value, legitimacy and relevance of science. Scholarship in the social studies of science goes further to argue that scientific knowledge itself is influenced by social and cultural factors, bringing the status of scientific knowledge as objective and neutral into question. This dissertation utilizes these two literature areas in order to derive an interactionist model of science-policy. A pragmatic framework is developed to focus on scientific processes in the policy sphere rather than on scientific content for addressing problems of science in policy. The proposition of this research is that processes that facilitate science-policy interaction contribute to the co-production of credible and legitimate policy relevant scientific information. The Intergovernmental Panel on Climate Change (IPCC) is used as a case study and is examined using this interactionist model. The IPCC was initiated in 1988 to interface with the international governmental community. The procedural adaptations that have taken place over the course of the past three IPCC iterations (1990, 1996, 2001) have increased governmental involvement in securing the legitimacy of the information produced and have indirectly led to considerable science-policy interaction. Two boundary processes, the review and approval processes, offered a sophisticated way for the user community (governments) to expose different and contradictory value frameworks while interrogating the underlying values and policy biases embedded in the science. The third boundary process, the policy relevant scientific questions, is viewed as a step in the direction of supporting an interactionist model of sciencepolicy. This research shows that science-policy interaction in the IPCC process has moved in the direction of a constructivist understanding without having been framed that way. Recommendations are made for ways to move from a focus on what is referred to here as weak co-production in the IPCC to the strong co-production of policy relevant scientific information in order to derive climate change information that is both authorized and imbued with meaning.  Table of Contents  Abstract  ii  Table of Contents  iii  List of Tables  vii  List of Figures  viii  List of Acronyms  ix  Acknowledgments  x  CHAPTER I: Conceptualizing a Pragmatic Approach for Scientific Advice in the International Policy Sphere . 1 1.1 Global Science in the International Sphere  1  1.2 The I P C C Case Study  3  1.3 Research Approach and Contribution  5  1.4 Structure of the Dissertation  6  CHAPTER II: The Hybrid Character of Science.  10  2.1 Philosophy of Science  10  2.2 Stages of Science Studies  11  2.3 Sociology of Scientific Knowledge (SSK) 2.3.1 Institutionalized Continuity in Science  .  13 16  2.4 Science as Practice 2.4.1 Science as Social, Conceptual and Material Hybrids  18 19 20  2.4.2 Science as Context-Dependent 2.5 Socio-Technical Hybrids  21  2.6 Implications of the Analysis  22  2.7 The Following Chapter  23  CHAPTER III: Bridging the Gap: Problematizing International Science 3.1 Models of Science in Policy  •  24 25  3.1.1 Science for Policy 3.1.2 Truth Speaks to Power 3.1.3 Politics of Science 3.2 Policy Relevant or Mandated Science 3.2.1 Problems with Translation 3.2.2 Value Frameworks in Mandated Science Box 3.1 The Multivalency of Science in Policy: The Case ofAlachlor in Canada  26 27 29 30 30 33 36  3.3 Scientific Process versus Scientific Content  37  3.4 Boundary Work, Boundary Objects and Boundary Organizations 3.4.1 Boundary Objects: Negotiating Representation  39 40 iii  41  3.4.2 Boundary Organizations: Managing the Science-Policy Interface 3.5 Boundaries in the International Sphere  42  3.6 Developing the Imbued Meaning Framework  44  46  Box 3.2 The Five Elements of the Imbued Meaning Framework 3.7 Implications of the Analysis  •• 48  3.8 The Following Chapter  49  CHAPTER IV: Methodology.  50  4.1 Particular setting  50  51  4.1.1 Why the IPCC was chosen 4.2 Case study  52  4.3 Using a Constructivist Approach  53  4.4 Research Methods  54  4.5 Sources of Data  54  55 56 58 64  Box 4.1 A Research Anecdote 4.5.1 Participant Observation 4.5.2 Interviews 4.5.3 Documents The use of documents in this research  65  66  Box 4.2 A Hypothetical Example of IPCC Inter-Textual Documents 4.6 Data Reduction and Analysis  ....67  67 69  4.6.1 Steps involved in reducing the data 4.6.2 The interpretation of data 4.7 Implications of this Analysis  72  4.8 The Following Chapter  72  CHAPTER V: Conventional Science-Policy? Understanding the Governmenta Presence in the IPCC. • 73 5.1 Including the User: Enrolling the Policy Network  74  5.2 History of the Science-Policy Interface  75  77 82  5.2.1 Development of the Intergovernmental Mechanism 5.2.2 Making Global Reports Relevant 5.3 Constructing the Second Assessment Report  83  5.3.1 Working Group Report Review Process 5.3.2 The SPM: The 'Translation' Process 5.3.3 The Synthesis Report  86 87 88  5.4 Constructing the Third Assessment Report (TAR)  5.4.1 The Policy Relevant Scientific Questions (PRSQ)  .  88  88  5.5 Inescapable Borderlands: Managing the Interface  91  5.6 Implications of Analysis  94  5.7 The Following Chapter  96  iv  CHAPTER VI: Science-Policy Interactions at the Summary for Policymakers and the Synthesis Report Interfaces 97 6.1 Participants at the Boundary  97  6.2 The Role of the Underlying Reports  99  6.3 Boundary Processes in the Summary for Policymakers and Synthesis Report  101  6.4 The Review Process  101  Table 6.1: Number of Review Comments in Examined Documents 6.4.1 The Bi-Directional Exchange of Materials and Discourses 6.4.2 Reflexive Feedbacks  102 103 105  6.5 The Approval Process 6.5.1 The Negotiation of Underlying Frameworks in the Approval Process 6.5.2 Negotiating Valid Information 6.5.3 Contact Groups 6.5.4 The Approval Process: A Negotiation of Words or Concepts?  106 108 109 113 117  6.6 The Policy Relevant Scientific Questions 6.6.1 Including the User in Framing What is Relevant 6.6.2 The Example of "Dangerous Anthropogenic Interference"  121 123 124  6.7 Authors' Varied Perceptions of the S P M and S Y R 6.7.1 Individual Perspectives 6.7.2 The Underlying Patterns in the Interviews 6.7.3 The Legitimate and Credible Combination  130 131 134 136  6.8 Out with the Idealized Model of Science-Policy, In with a Pragmatic Model CHAPTER VII: Concluding Remarks and Recommendations  138 140  7.1 Constructivist vs. Unidirectional  141  7.2 The I P C C as Process: Relevant Enough and Neutral Enough  142  7.3 Interactive Boundary Processes  143  7.4 Perceived Models of Interaction  145  7.5 Imbued Meaning: A Pragmatic Approach to Making Co-Production Transparent 145 7.6 Negotiating Credibility  148  7.5 Responsive Interaction  149  Bibliography Appendices.  •  151 •  169  Appendix A : Questions for the IPCC T A R Meeting in Accra, Ghana  169  Appendix B : Informal I P C C Documents Examined  171  Appendix C : Procedural Adaptations in the SAR and T A R The SAR SPM: Policy Relevant Scientific Information The TAR SYR: Question Driven Approach  174 174 176  Appendix D: I P C C as Mandated Science: A l l products are negotiated  178  The Bastion of Economists The Knowledge Net There is No Such Thing as Neutral Information  :  178 180 182  Appendix E : Authors' Perceptions of the SPM and SYR 1) View the SPM (and SYR) as a government-approved translation.... 2) View the SPM (and SYR) as a negotiated hybrid accompanied by trade-offs 3) View the SPM (and SYR) as a political government document  183 183 183 184  Appendix F: List of Dissertation Related Publications and Presentations  185  vi  List of Tables  Table 4.1: List of Authors, Governmental Officials and Observers Formally Interviewed  62  Table 4.2: The Advantages and Disadvantages of Methods Used  70  Table 5.1: Differences in Structure Between the F A R and the SAR  74  Table 5.2: A Historical and Causal Typology for the Increased Science-Policy Interface in the IPCC 90 Table 6.1: Number of Review Comments in Examined Documents 102 Table 6.2: Number of Participants in the WGIII SPM Approval Process  107  List of Figures  Figure 3.1: Three Models of Science-Policy  51  Figure 5.1: Structure of IPCC Reports  92  Figure 6.1: Climate change - an integrated framework  Ill  Figure B . l (Appendix): The Negotiation of the SPM  175  Figure B.2 (Appendix): The Framing and Negotiation of the SYR  177  Figure D. 1 (Appendix): The Evolution of the IPCC Assessment Process  180  viii  List of Acronyms  AGGG  Advisory Group on Greenhouse Gases  AOSIS  Alliance of Small Island States  CDM  Clean Development Mechanism  COP  Conference of the Parties  DES  Development, Equity and Sustainability  ENGO  Environmental Non-Governmental Organization  FAR  First Assessment Report (to the IPCC)  4AR  Fourth Assessment Report  FOD  First Order Draft  GARP  Global Atmospheric Research Program  GCM  General Circulation Model  GHG  Greenhouse Gases  IGBP  International Geosphere Biosphere Program  INC  Intergovernmental Negotiating Committee (for the Climate Convention)  IPCC  Intergovernmental Panel on Climate Change  JI  Joint Implementation  NAS  National Academy of Science  NGO  Non-Governmental Organization  OECD  Organization for Economic Cooperation and Development  PRSQ  Policy Relevant Scientific Questions  SAR  Second Assessment Report (to the IPCC)  SBSTA  Subsidiary Body on Scientific and Technical Advice  SOD  Second Order Draft  SPM  Summary for Policymakers  SRES  Special Report on Emissions Scenarios  SYR  Synthesis Report (to the IPCC)  TAR  Third Assessment Report (to the IPCC)  TOD  Third Order Draft  TSU •  Technical Support Unit  UNEP  United Nations Environment Programme  UNFCCC  United Nations Framework Convention on Climate Change  WGI, II, III  Working Group I, II, III  WMO  World Meteorological Organization  Acknowledgments Yes Noah, we can save the polar bears! This research is dedicated to my 4-year old nephew and his little brother Aydan. This research project was sponsored in part by Environment Canada's Climate Change Action Fund (CCAF). This generous support provided me with the opportunity to travel to Accra, Ghana, Washington, D C and London, U K to observe the IPCC processes and to perform interviews with IPCC scientists during what I fondly refer to as the "glory days". There are many people to acknowledge for the encouragement and support particularly in the post-glory days. This begins with the intellectual support of my four-committee members, each with varying degrees of intrigue and participation in the project. First and foremost, this project came into existence through the inspiration and support of my research supervisor, Dr. John Robinson. His ability to motivate, inspire, think and challenge has contributed to an incredible learning throughout this process. I am indebted to his unrelenting faith in my abilities and vision in the project. Dr. James Tansey for offering incredible guidance throughout the project and for understanding the person behind the PhD student. Dr. Les Lavkulich for always having a door open and for always being there to offer support. Dr. Douw Steyn for taking an interest in the subject matter and for providing a thorough and interesting perspective. I would also like to acknowledge the late Dr. Stephen Straker for his contributions. Our sporadic threehour lunches stimulated my thinking and helped guide me through the social studies of science literature. It is necessary to acknowledge my colleagues and the staff at the Sustainable Development Research Initiative (SDRI) who over the past two years have generated a lively agora in the building, always keeping me connected to a larger picture. In particular thank you to Misty for her critical eyes, which served as a tremendous help during the writing phase. To my officemates Soma and Estelle, for sharing this process as a communal experience and for the sticky note that hangs as a motivation in the office disclosing the gloomy completion statistics of female doctoral students in Canada. To my colleagues at the Resource Management Environmental Studies (RMES) program whose diverse and intriguing interests contributed to an exemplary interdisciplinary experience. I am eternally grateful for the love, support and understanding from friends and family for their seemingly clandestine efforts to prevent this dissertation from becoming another Canadian statistic. In particular thank you for commiserating Arne, Rachel and Raul, and for the regeneration and renewal Bronwen and Marcia. Thank you to my siblings David, Meghan and Jaime for their wellspring of faith. I am eternally thankful to my mother, Virginia, for her encouragement, creativity and for sharing her oneline epiphanies inspired by my research and to my late father, who has influenced this work in incredibly nuanced ways. A n incommunicable appreciation goes to Alistair, who celebrated my smallest achievements and helped me through the hardest times; for this everyday support, patience and determination I am so grateful.  x  CHAPTER I: Conceptualizing a Pragmatic Approach for Scientific Advice in the International Policy Sphere For decades climate and atmospheric scientists have been monitoring the atmosphere and the climate and have recently begun to observe signs of considerable changes to the climate system (Ramanthan et al. 1985; N R C 1983; Arrhenius 1997). Scientific conferences such as the World Climate Conference of 1979 were points of congregation for scientists working on climate issues from different disciplinary backgrounds focusing independently on atmospheric, oceanic or terrestrial systems. Despite this considerable scientific activity, the issue of climate change did not make political waves until the late 1980's with the establishment of the Intergovernmental Panel on Climate Change (IPCC). Emerging from the successes of the Vienna Convention and the ratification of the 1987 Montreal Protocol, the IPCC was established as a scientific assessment body capable of producing relevant scientific information that could then feed into the international policy sphere. The IPCC is the scientific assessment body that compiles, assesses, and interprets scientific information pertaining to (i) climate science in working group I (WGI), (ii) the biophysical and social impacts, adaptations and vulnerabilities associated with climate change in working group II (WGII) and (iii) possibilities for mitigation in working group III (WGIII).  As a global commons problem, climate change raises profound questions about how to deal with such a vast and truly borderless problem at the international scale. It also provides a particularly powerful case study through which to examine the role science plays in decision-making.  1.1 Global Science in the International Sphere The IPCC scientific assessment was co-initiated by the World Meteorological Organization (WMO) and the United Nations Environment Programme (UNEP) as a response to significant international concern about climate change. The IPCC was initially set up as a one-time international assessment, a collective of established scientists, many of whom, had been involved in the successful ozone assessment (Boehmer-Christiansen 1994a,; Agrawala 1998). These scientists undertook a comprehensive effort to evaluate, select, collate and interpret peer-reviewed climate literature on the science, impacts, and response strategies. Many of these scientists were also involved in constructing large-scale climate models and general circulation models (GCM) in national and international scientific programs such as the International Geosphere Biosphere Program (IGBP) and international centers devoted to investigating climate science such as the Hadley Centre and the National Center on Atmospheric Research (NCAR). Such models were used to project potential ecological and socio-economic impacts of climate changes at different greenhouse gas (GHG) concentrations in the atmosphere and based on different degrees of 1  warming, could then be used to speculate on potential response options. The assessment was intended to provide a state of the art climate report for what was then a small international policy audience. However, over the past 17 years and over three iterations of the science assessment process, the IPCC has expanded to include national governments, which are members of the W M O and/or the UNEP, as an audience for the 11 Climate Change reports and the 5 Special Reports. Also included in the process are 1  observers or "witnesses" from civic and industry organizations such as the environmental non2  governmental organizations (NGO) Greenpeace and the industry N G O , the Climate Change Coalition for instance. In short, the IPCC has become the most elaborate and extensive scientific assessment ever undertaken involving over hundreds of scientists, including external reviewers, 120-140 national governments, and 20-30 observer organizations. IPCC findings provide a consensual and policy relevant understanding of the climate problem considered necessary to facilitate international cooperation, however the climate issue has proven to be much more contentious than the ozone issue ever was.  The IPCC is intended to provide credible, transparent and objective reports combined with policy relevant but not policy prescriptive summaries to be diffused to the international policy sphere. A recent IPCC pamphlet outlined its chief mandate:  Human activities now occur on a scale that is starting to interfere with complex natural systems such as the global climate. Many people argue that climate change poses a risk of serious or irreversible damage and that the lack of full scientific certainty should not be used to postpone action. Others believe that the risk is manageable and does not justify a major shift in how financial and human resources are invested. Because climate change is such a complex and challenging issue, policymakers need an objective source of information about the causes of climate change, its potential environmental and socioeconomic impacts and possible response options (IPCC 2004).  In order to fulfill these objectives, the IPCC has increasingly included the user community or governments in the review and acceptance of the underlying working group reports, and in reviewing and approving their respective summaries for policymakers on a line-by-line basis.  Despite this massive undertaking by the international environmental science community there have been many debates and controversies regarding the credibility and accuracy of the science surrounding climate  The Special Reports include the Regional Impacts of Climate Change (1997), Aviation and the Global Atmosphere (1999), Methodological and Technological Aspects of Technology Transfer (2000), Emissions Scenarios (2000), Land Use, Land Use Change and Forestry (2000). Edwards and Schneider (1997) refer to the civic institutions with observer status in the IPCC "witnesses". Thenpresence bear witness to the legitimacy of the processes that occur in this forum. 1  2  2  change.  3  Even today, though the Kyoto Protocol, an international agreement on climate change has now  come into force,, both scientific and political sceptics argue that there is no climate problem. For instance, in the US Senate, decision-makers such as Senator James Inhofe of Oklahoma, the Chairman of the US Senate Committee on Environment and Public Works argue that, "much of the debate over global warming is predicated on fear, rather than science" (Inhofe, 2005). In the 10 round of the international th  climate negotiations in Buenos Aires, the US delegation remained unwilling to participate in the internationally binding Kyoto Protocol under the United Nations Framework Convention on Climate Change (UNFCCC). The leader of the US delegation, Paula Dobriansky, commented on the unsatisfactory state of climate science stating, "Science tells us that we cannot say with any certainty what constitutes a dangerous level of warming and therefore what level must be avoided" (Inhofe, 2005).  4  Sceptical attitudes still exist in science, politics, industry, and the public, despite significant assessment efforts on the part of the IPCC and other scientific bodies (Baliunas 1999, Singer 2001, Lindzen 2001).  5  The implications of addressing and not addressing climate change are profound and raise a particularly powerful way to question the role that science plays in decision-making.  1.2 The IPCC Case Study  The IPCC assessment establishes the scientific basis and understanding of climate change and has undergone reappraisal roughly every five years (1990, 1996, 2001) since 1990. It is comprised of its own set of practices, norms, values and actors that, given the multitudes of international studies and data, are charged with the task of achieving consensus on the state of the climate and its changes. As will be shown in the following dissertation, in response to some of the criticisms over the past decade which have brought climate science, its methods and the IPCC process into question, the IPCC process has repeatedly changed and adapted its procedures to improve both the perceived credibility and legitimacy of the information produced. . 6  v  In 2000, US President Bush opted out of the ratification of the Kyoto Protocol (1997) under the United UNFCCC. His argument was that the implications of the collective action drafted under the international agreement would be too costly for the USA. In a sense, this is a profound admission and shift in response. It doesn't argue that the science doesn't support climate change; rather it says that it can't describe what constitutes 'dangerous' climate change. See the Science and Environmental Policy Project (SEPP) website for an impressive list of contrarian perspectives on climate change recently published in North American press. http://www.sepp.org/NewSEPP/newsepp.html. There are a number of studies that argue that popularized scepticism toward climate change is less to do with significant controversy and divergence in the scientific literature (Oreskes 2004) and has more to do with the shift from scientific discourse of the IPCC to more popularized discourses in the media that are veiled under notions of balance in journalism (Boykoff and Boykoff 2004). For the purposes here, credibility refers to the scientific adequacy of technical evidence and arguments; legitimacy refers to the perception among participants that the information has considered divergent values and beliefs, has 3  4  5  6  .3  The IPCC assessment has two innovative products created to include the policy community in negotiating the representation of the policy relevant scientific information: the Summary for Policymakers (SPM) and the Synthesis Report (SYR). Each of the Working Group (WG) reports is accompanied by a SPM similar to an executive summary. The second assessment report (SAR) and the third assessment report (TAR) produced a fourth report, the Synthesis Report (SYR), that provides an overview of the findings from the three working groups. The SPM and SYR reports are the most concise, accessible and thus useable products of the EPCC; as their titles suggest they provide a summary and synthesis of W G findings that is accessible and relevant for policymakers in an effort to generate policy relevant scientific information.  The SPM and SYR, formally described as translation processes (better ways to communicate scientific findings to the political sphere), are intended to increase the policy relevance of the scientific information produced by the Working Groups, while expediting its transfer to the international policy sphere (Bolin 1994; Agrawala 1998; van der Sluijs 1997). The SPM is defined as a "component of a Report, such as an Assessment, Special or Synthesis Report, which provides a policy  relevant but policy neutral summary of  that Report" [emphasis added] (IPCC 1999). The SYR is defined as a report "that synthesizes and integrates materials contained within the Assessment Reports and Special Reports and [is] to be written in a non-technical style suitable for policymakers and address a broad range of policy  relevant but policy  neutral questions'" [emphasis added] (IPCC 1999). The SPM and SYR instruments ensure that information crosses into and is consensually approved in the governmental sphere. This interface is formally managed; experts and governments conduct a formal review and the revised draft is approved line-by-line (in the case of the SPM) and is adopted section-by-section (in the case of the SYR) by national governments. In other words the policy relevant products of the IPCC have consensual "buy-in" from national governments. In its effort to derive policy relevant but policy neutral findings through the SPM and SYR, the IPCC represents an ideal case study of the science-policy interface and the ways that scientific information is diffused and interpreted in the international policy arena.  What is of interest here are not the final SPM and SYR products which have been examined and analyzed extensively, but instead both the formal and informal processes that contribute to the seemingly paradoxical requirement of being policy relevant while remaining policy neutral that also, lead to a seemingly impossible task of achieving governmental approval, adoption and thus "buy-in" to the final products. The main research question is: What are the formal and informal processes and protocols in the  been unbiased in its conduct and fair in its treatment of opposing views and interests; and salience refers to the relevance of the assessment to the needs of the decision-makers (Clark et al. 1997).  4  IPCC that contribute to the production of policy relevant scientific information that remains credible and legitimate?  1.3 Research Approach and Contribution  There are three goals in this dissertation. The first is to contribute to a literature about the nature of science in society. Scholarship in the social studies of science (SSS) utilizes a constructivist approach in order to examine both the formal and informal processes and practices involved in scientific claim making (Latour 1987; Bloor 1996). A constructivist approach makes a formal distinction between the theory and rhetorical descriptions of science, the idealized culture of science (Latour 1987, 1993; Gieryn 1989) and the actual practice of science (Pickering 1992; Clarke and Fujimura 1992). SSS researchers argue that science is not decontextualized from, but rather is immersed in, the social, political and cultural environment (Shapin et al. 1985; Latour 1987, 1988, 1993; Barnes et al. 1996; Knorr-Cetina 1981, 1992, 1999; Longino 1990, 2002). The argument continues that scientific claims are therefore not unitary and monovalent descriptions of the natural world "out there" but are instead context dependent: idiosyncratic practices, social assumptions and cultural commitments are interwoven in the representation of scientific claims. With SSS interpretations of science, the conventional understanding that science and policy are separate spheres of activity and discrete professional communities is brought into question. Once social processes and cultural commitments are identified in underlying scientific claims-making activities it becomes clear that the social and scientific spheres are not separate in practice but are instead interpenetrating forces that influence and mutually legitimate one another (Latour 1987, 1993; Callon and Latour 1992). The SSS literature is important for understanding the status of scientific knowledge itself and becomes critical in examining the role science plays in policy. SSS literature undermines the notion of 'truth' thereby bringing the conventional approach of science for policy or 'truth speaks to power' (Price 1965) into question. This conventional approach continues to dominate theoretical understandings despite considerable literature that highlights its failures (Salter 1988; Jasanoff 1995, 1997, 2003). This dissertation will attempt to provide alternative theoretical insights into the status of science and its use in the international policy sphere.  The second goal of this dissertation is to contribute to a theoretical debate but also to utilize SSS approaches to address a real problem. The assumption here is that what is necessary to understand and respond to global environmental problems is policy relevant scientific information that is credible and legitimate to both science and policy communities. Scientists are relied on in increasingly technical policy problems to identify and describe problems and potential hazards, while policymakers (and policy processes) determine what information is policy relevant, and contribute to politically feasible and  5  appropriate responses. As the term suggests, both communities are required to produce policy relevant scientific information that remains credible and legitimate to both communities. In this dissertation a framework is developed that proposes a pragmatic use O f the social studies of science information in order to encourage science-policy interaction and to contribute to the production of both reliable and meaningful policy relevant scientific information. 7  The third goal is to bring the IPCC under critical examination. The goal is not to determine how the scientific claims of climate change have been and are being constructed (other researchers have done this) (Boehmer-Christiansen 1994a,b; Shackley and Wynne 1997; Shackley et al. 1998; van der Sluijs et al. 1998). Instead the goal is to analyze and evaluate ways in which policy relevant scientific interpretation is selected, represented and disseminated in the science-policy interfaces of the SPM and SYR. SSS research of climate change issues has focussed on the contents of the science itself (Shackley et al. 1998, van der Sluijs et al. 1998), while others have focussed on the evolution of the institutional structures contributing to the IPCC assessment and its interface with policy (Boehmer-Christiansen 1994a,b, Shackley and Wynne 1995, 1997). However there has not been SSS research conducted that explicitly examines the processes of science-policy interaction in the context of the IPCC assessment. Sciencepolicy interaction has been identified as an under-estimated element of global assessment design (Farrell et al. 2001). In order to explore the science and policy relationship in the IPCC, it necessary to examine the underlying processes in the SPM and SYR that contribute to the production and diffusion of "policy relevant and policy neutral scientific information" (IPCC 1999). These science-policy interfaces also provide innovative areas to test some of the theoretically-based findings introduced in the previous two sections.  1.4 Structure of the Dissertation One cannot approach an environmental problem without recognizing the reinforcing interactions and interrelationships between the natural and social domains. Nor can one conceive of environmental problems independent of ethical questions relating to equity, the moral dimensions of humans' role in the world and principles of democracy where governments must consider the desires and best interests of its citizens. Yet 'modern' institutions continue to describe science as rational, value-neutral and objective descriptions of the natural world 'out there' distinct and autonomous from the socio-political landscape. This poses a problem for environmental issues where the natural and social intermingle and reinforce in dynamic ways. A new understanding of the term 'science' has been advanced in recent scholarship and  Reliable refers to information that is certified or ratified as being the best collective judgement by the scientific and broader user communities through 'due process'. Meaningful refers to information that is framed by both the scientific and user communities and is reappraised through time. 7  for the purposes here, refers to the materials, theories, instruments, ideas, discourses, values, institutions and actors that construct understandings of objects and phenomena that surround us. This understanding of science involves an investigation into the situated and contingent practices that facilitate, shape and stabilize particular theoretical and empirical understandings of the world (Wynne 1992, 1996; Jasanoff and Wynne 1998; Jasanoff 2003, 2004). It involves investigating science in practice and the ways science is both shaped by and shapes the social and cultural sphere. Chapter 2 examines the social studies of science scholarship and suggests that science is necessarily a human product that is made up of particular cognitive, social and political processes.  Chapter 3 explores the problems associated with the continual shift of cognitive authority between scientists and politicians in technical decision-making processes. Where scientific consensus is contested and uncertainties highlighted, scientists lose public and political authority to contribute to public policy decisions. This leads to a backlash where authority relevant to a problem and its informational requirements are transferred from science back to the political sphere where, in general, calls for more research to reduce scientific uncertainties are made. In environmental policy this becomes especially problematic. The complexity of environmental issues results in the use of different disciplinary approaches that emphasize different aspects of the issue and thus on the adequacy and sufficiency of the information that typically leads to scientific controversy. This controversy or lack of consensus thereby legitimizes policy inaction on critical and potentially urgent issues. Chapter 3 will examine the potential for a middle ground between idealized 'truth speaks to power' and highly relativistic 'power speaks to truth' conceptions of the relationship between the scientific and policy communities.  The Imbued Meaning framework developed in this chapter is based on research in the social studies of science and on lessons from constructivist accounts of science-policy interaction. It proposes five procedural innovations (these are neither comprehensive nor exhaustive) intended to improve the management of science-policy interactions. The goal is to derive a pragmatic framework that manages science-policy interaction to contribute to the production of policy relevant scientific information maintaining the perceptions of credibility and legitimacy among the participating science and policy communities (as well as non-participating professional communities).  Chapter 4 describes the IPCC as a case study. Constructivist case-study approaches have been used to investigate the processes involved in construction of climate science (Shackley and Wynne 1995, Shackley et al. 1998, van der Sluijs et al. 1998) and have been used to trace the history of the EPCC process (Agrawala 1998a,b; Boehmer-Christiansen 1994a,b). However, a constructivist account of the processes and protocols internal to the IPCC's SPM and SYR has not been developed. The SPM and  7  SYR represent two science-policy interfaces where scientists and policymakers are brought together to approve particular representations of what are the most-cited policy relevant scientific documents on climate change. How are these documents derived? What are the processes that contribute to this common representation of the scientific information?  Access was gained to three IPCC meetings. The first was the 17 Session of the WGIII Session of the th  Panel held in Accra, Ghana in March 2001. This meeting focused on the acceptance of the underlying report from WGIII in the T A R and the approval of the WGIII SPM. The second was at the scientific meeting held in Annapolis, Washington in June 2001. This was a meeting of coordinating lead authors (CLA) involved in drafting the SYR report and its SPM. The third was the 21 Session of the Panel held st  in Wembley, U K where the SYR and its SPM were accepted and approved by delegations. Document analysis, participant observation, and participant interviews are used to examine: 1) the science-policy interface in the IPCC, 2) the science-policy interactions in the SPM and the SYR, and 3) the ways scientific participants view the process. Understanding the formal and informal processes that contribute to this common representation of information provides insights into ways that science and policy communities can interact to derive a common representation of scientific information while remaining credible, legitimate and relevant.  Chapter 5 provides a brief description of the origins and historical developments of the IPCC process. This account reveals the ways that the boundaries between the science and policy communities emerged and the ways that the science-policy interface was formally constructed and managed. This chapter follows the iterations of the IPCC assessment reports to examine the ways the IPCC adapted its procedures around external criticisms and honed its credibility among scientists as well as its legitimacy among governments. This elasticity in the process and the particular objectives of the IPCC to translate climate science into the policy sphere, led to the somewhat ad hoc introduction of two instruments, the SPM and SYR. These instruments were intended to separate and purify the boundaries between the science of the underlying reports and the policy documents for the governments while simultaneously transferring policy relevant but policy neutral scientific information to the international policy sphere.  Chapter 6 shows that on closer examination the 'translation' processes involved in producing the policy relevant documents of the SPM and SYR, were instead the products of a complex series of practices and interactions that amounted to something other than a unidirectional translation of policy relevant findings. The review and approval processes represented areas of formal and informal interaction. The latter involved a formal negotiation between C L A ' s and national governments over what information is  8  considered to be consistent, balanced and policy neutral. There are formal procedures that manage interaction in this forum in the review process, the approval process and the policy relevant scientific questions in the SYR that establish processes that contribute to the production of policy relevant scientific information. However I also observed the informal processes and protocols emergent from these interactions that contribute to the successful line-by-line approval of the SPM (or the section-by-section adoption in the case of the SYR). It is within these informal 'boundary processes' that discursive interactions occurred between scientists and governmental officials about what is considered to be a consistent, balanced and policy neutral representation of the underlying W G reports.  9  CHAPTER II: The Hybrid Character of Science Scientific knowledge has come under considerable scrutiny in scholarly works over the past four decades (Bloor 1976; Knorr-Cetina 1981; Shapin et al. 1985; Ashmore 1989; Pickering 1992; Clarke and Fujimura 1992). The disciplines of history, sociology, philosophy, and psychology have increasingly brought scientific knowledge under examination to determine how and whether scientific methods enable humans to discover a transcendental and unitary natural world apart from their social and cultural contexts. This research was mobilized to better understand why one scientific claim comes to dominate over another; why one is believed more than another. This scholarship is therefore critical for examining the role of science in policy that has conventionally been conceived of as 'truth speaks to power'. In order to examine this relationship it is first necessary to examine the status of scientific knowledge itself.  There are two well-established ways of looking at the role and status of scientific knowledge: the first focuses on philosophical issues concerning the epistemological and ontological status of the knowledge claims of science, and the second focuses on the social contexts and practice-settings in which such claims get made. While there is considerable overlap between the two approaches, at the extremes they suggest either a purely rationalist or a purely relativist account of science. The purpose of this chapter is to explore the degree to which recent findings in the social studies of science (SSS) can inform our understanding of how to think about the role of science in society. A concern among scientists and users of science that a constructivist account necessarily undermines the authority and meaning of science itself will be examined and challenged. In the following two chapters a case will be developed for an 8  increasingly pragmatic role for constructivist accounts of 'science in action' (Latour 1987).  2.1 Philosophy of Science  No one denies that the discovery and validation of scientific facts is a fundamentally human activity. What is in dispute however is the potential of this human activity to transcend social and cultural confines to access some natural and transcendental world that exists beyond humans. Some influential approaches in the philosophy of science have focussed on constitutive values and norms in science that attempt to achieve methods and practices that ensure the cognitive autonomy and objectivity of scientific knowledge (Longino 1990, 2002; Rouse 1996). These positivist approaches have derived theories and methods that  The epistemological struggles introduced by examining the taken-for-granted aspects of science has contributed to persisting friction between the cultures of the natural and social sciences. Some SSK practitioners belong to a strong relativist camp that believe that if physics can be explained by sociology then the hierarchy of the sciences can be inverted (discussed in Ashmore 1989; also see Callon and Latour 1992). 8  10  constrain the influence of contextual values in discovering the truths in nature (Popper's theory of falsification for instance). The authority of science is primarily based on the constitutive claims of objectivity, autonomy and value-neutrality within the scientific enterprise (Longino 1990). However, the SSS literature argues that science examined as a social process highlights the underlying normative and value-laden commitments involved in the scientific endeavour that bring objectivity, autonomy and neutrality into question (Collins 1981d, 1982; Collins and Pinch 1998; Bloor 1991; Pickering 1992; Callon and Latour 1992; Barnes and Bloor 1996). Rather than science simply decoding the universe 'out there', SSS researchers suggest that cognitive, ideological, behavioural and social factors influence what phenomenon, objects, or system is examined, how it is examined and the ways interpretations are made (Ashmore 1989; Latour 1987, 1993; Clarke and Fujimura 1992; Pickering 1992). It is through these investigations that discovery and rationality in the content of scientific knowledge is challenged. The contingency of claims are considered to be based, to some degree, on the goals of the practitioner and of the discipline (both theoretical and professional) (see Latour and Woolgar 1979; Latour 1987; Shapin, Schaffer, and Hobbes 1985; Collins and Pinch 1982; Knorr-Cetina 1982), the differences in the practices and resource availability of labs (tools, machines, technologies employed) (see Hacking 1992, KnorrCetina 1992), and the socio-political and cultural context from/for which information is derived (relating to funding sources, instrumentality, cultural, social and political goals and needs) (see Latour 1987; Yearley 1996; Wynne 1982, 1989b, 1992, 1992a; Clarke and Fujimura 1992; Jasanoff and Wynne 1998; , Shackley et al. 1998). SSS finds that the social context and that existing conceptual schemes are an intrinsic part of the cognitive landscape and thus intrinsic and extrinsic influences on the process of scientific claim making (Barnes, Bloor et al. 1996; Longino 1990, 2002). Rather than seeing these processes as extrinsic and attempting to 'purify' them from scientific content (Latour 1987, 1993; Gieryn 1983, 1999), it may be more desirable to examine and analyze both the constitutive practices of science and the contextual factors that contribute to how it is framed and shaped. It has been argued (Jasanoff and Wynne 1998) and is argued here that the social and contextual factors that influence the way we construct the world need to be transparently considered in order to understand science in a realistic rather than idealistic or rhetorical way. This research compels the social researcher to examine the judgments, choices and factors that shape knowledge claims or, in the words of Latour (1987), to analyze 'science in action'.  2.2 Stages of Science Studies  The sociology of science emerged from the understanding that contextual values can influence the direction and interpretation of scientific knowledge (racist ideology and gender bias in science were identified as examples of this) (Haraway 1991; Harding 2000). Robert Merton recognized that the products of science do not emerge from nature but from people (in Ziman 1968). Merton analysed the 11  ways ideology, bias and interests are situated within the scientific endeavour. His conclusion was that efforts were needed to monitor the work of individual scientists and to further standardize scientific practices. The Mertonian norms were based on a rational actor approach to object/subject relations 9  suggesting that cognitive practices could be distinct from the social landscape if appropriate rules were followed.  10  This was considered to be insufficient to address the complex social practices and  interactions in science.  11  The sociology of science was radicalized by the publication of Kuhn's (1962) work  The Structure of  Scientific Revolutions. This work revealed the way that the theoretical context influenced both the selection of scientific questions asked based on 'doable' problems and pre-existing tools and practices maintained a particular direction in 'normal' scientific activities until the puzzle-solving methods and capabilities were brought into question. At which point a whole new scientific paradigm can be initiated. Kuhn's work revealed the ways that scientific work is stabilized within disciplines of practice that utilize pre-existing tools and discourses to ensure that normal puzzle-solving capabilities 'fit' within existing theoretical structures. It revealed the ways that the 'rules' for thinking about the universe and assessing evidence and proofs is based on existing conceptual frames that can become obsolete at some point in the future, initiating a scientific revolution (Yearley 1994). Kuhn demonstrated how scientific facts at any given historical moment both reflect and impact on the wider social contexts in which they were produced (Latour and Woolgar 1986; Latour 1987; Shapin et al.1985).  The challenge for SSS then, was to move beyond the analysis of truth in nature and to critically examine the social processes and practices that underline the technical content of scientific knowledge instead. This has been referred to as opening the "black box" of scientific fact construction (Latour 1987). Whereas the philosophy and sociology of science have attempted to demonstrate science as a positivist epistemology capable of establishing truth, SSK reveals the ways that the social cannot be eliminated from cognitive practices (Latour 1987; Longino 1990, 2002). In the 1970's, a flurry occurred in the development of research programmes such as Bloor's Strong Programme (sociology of scientific knowledge (SSK)) (Bloor 1991; Barnes et al. 1996; Wynne 1982), the Paris Group (socio-technical hybridity) (Latour and Woolgar 1979) and the Tremont Group (symbolic interactionism) (Star and  The Mertonian norms are communalism, universalism, disinterestedness, and organized scepticism (CUDOS). This relates to an ongoing debate in contemporary sociology; is it safe to conclude that actors follow rules when they are instituted? Answering yes focuses on aggregating overall performance and assumes that actors are rational, that they consider their actions in a deliberate way. Answering no assumes that actors are heterogeneous operating with different and varied interests. The illegal economy is an example of the latter, where agents are not operating in rational ways or according to prescriptive rales. As Ashmore (1989) notes the sociology of science failed primarily to overcome problems with reflexivity. Sociology of science treated science as unalterable and immune to self-criticism while utilizing the norms and practices from the natural sciences.  9  10  11  12  Griesemer 1989; Clarke and Fujimura 1992; Fujimura 1989, 1992). Common to all was a concerted effort to understand the underlying assumptions, judgments and social processes, that made-up the foundation of scientific inquiry bringing the technical content of science under a sociological lens.  2.3 Sociology of Scientific Knowledge (SSK) With the emergence of Bloor's Strong (Relativism) Programme in the 1970's, a new research area was initiated called the sociology of scientific knowledge (SSK) that challenged the notion that scientists are rational agents who comply with rules and norms. SSK scholars suggest that contextual values and influences cannot be expunged from scientific research because values, beliefs, and norms of the sociopolitical sphere influence the cognitive frameworks of the practitioners themselves (Bloor 1976; Ashmore 1989). Research in the social studies of scientific knowledge use constructivist approaches that question the notion of a monovalent and unitary understanding of science (Wynne 1982, 1989b, 1992, 1992a). The hegemony of theory and its especially privileged position in shaping scientific practice have become problematized under these approaches (Clarke and Fujimura 1992; Pickering 1992). The empirical work in the social studies of science reveals that theories, ideas, research and justification do not float freely but are instead its persuasive value is negotiated among 'certified' specialists for the ways that it 'fits' within the underlying values, discourses, and theories of particular disciplines; disciplines that have common theoretical frameworks and methods.  In this approach the culture of science is viewed as heterogeneous and is highlighted as neither unitary nor consensual, but as the craftwork of actors, instruments, disciplinary, social and cultural commitments negotiated in particular contexts (Latour 1987). It is posited that scientists have different ideas, commitments, and disciplinary lenses through which to pursue their research. SSK scholars argue that both professional and contextual influences have bearing on which questions are asked, which theories and methods are chosen and how data are interpreted (Bloor 1991; Barnes et al. 1996). Laboratory studies reveal the ways different theories, methods, instruments and 'tinkering' lead to different results and claims about a particular phenomenon (Latour and Woolgar 1979; Latour 1987; Knorr-Cetina 1992; Nader 1996). These variations lead to different weights given to various kinds of evidence being presented. Practitioners may hold different opinions about the reliability of types of equipment used, differ in estimations about the reliability of colleagues' work and find greater merit in their own direction (Yearley 1994). Analyzing science from a sociological perspective became a way to challenge the discursive authority of science that claims that scientific methods are the result of discovery and verification (or 'proof) of objects and phenomenon in the natural world. Social studies of science reveal the ways that interpretive flexibility and contestation of theories and data involve micro-negotiations over the value of the observations and interpretations as well as the legitimacy accorded to outcomes or 13  processes of validation (Collins 1985). Micro-negotiations were considered to be a way that scientific controversy moves from disordered heterogeneity to ordered stability in practice.  The ways that scientists certify results and validate hypotheses have been shown to be less unitary and more ambiguous than often exhibited in scientific discourse. Research into the methods of validation and verification reveal that scientific experiments often do not directly access scientific facts but instead knowledge is formed through the negotiations among a "core-set" group of scientists who come to agree on which data and methods are sufficient to make a particular claim. For instance, Collins' (1985) analyzed the micro-negotiations of physicists brought together to validate particular research and interpretations of gravity waves. He highlights a sphere of micro-politics and negotiation rather than a methodological standard that led to the validation of a scientific claim. By doing so, he reveals ambiguity in the final interpretation; validity was negotiated behind-the-scenes among a disciplinary 'core-set'.  12  Longino (2002) notes that the progress of science is dependent on the notion of continual extension of scientific knowledge. In earlier work, Longino (1990) suggests that validation of theories cannot be consistently tested within science. Instead a "systematic and unifying treatment of phenomena enables us to interact with the natural world with reliable expectations. A methodology that legitimates the stabilization of inquiry thus serves some constitutive ends of knowledge seeking" (Longino 1990). The stabilization of knowledge often occurs with the subordination of rigorous and critical reflexivity (Ashmore 1989) and validation in order for scientific information to accumulate and develop a reservoir of understanding. This subordination is considered necessaiy "to avoid the endless testing and constant generation of new explanatory frameworks that would subvert knowledge extension..." (Longino 1990, 9). Similar propositions have been made that reveal the ways that normal scientific inquiry must 'fit' with understanding in order to advance scientific research (Latour 1987, 1993; Downey 1992).  In order for scientific activity to proceed on an appearance of progress, trust among scientific practitioners is essential (Longino 1990). Scientists cannot analyze and replicate every detail of every experiment. Therefore colleagues must trust the abilities and rigour of the methods of science such as peer-review and reputations of one another. This trust is brought into question when the sufficiency, adequacy of methods and interpretations are evaluated differently among different researchers. Where scientific dissent exists, power relations and social networks among competing scientists have been shown to become critical in the stabilization of one claim over another (Latour 1987; Longino 1990). For  Collins' (1985) has used a similar line of argumentation to challenge the replication and validation process. Stating that experiments cannot be exactly reproduced in different laboratories with different skills and interests, Collins' determined that validation processes are the result of 'core-set' micro-negotiations. He asserts that these tests of tests or replications of replications leads to what he called 'experimenter's regress'. 12  14  example using historiographical analysis, Latour (1988) posits that Pasteur's theory of bacterial growth, which sought to explain how life existed in oxygen-free environments, dominated over Pouchet's theory of spontaneous generation, due to the networks and alliances that Pasteur was able to access and enrol. Latour (1988) followed the history of the science to suggest that the reasons Pasteur's claim became dominant was more to do with external variables such as reputation, the ability to mobilize resources, and networks that facilitated his ability to enrol allies in his particular theory. Similarly, Longino (2002) analyses the theory of'panselection' in biology and argues that the synthesis of Darwin's natural selection and Mendelian genetics was as much about discouraging Larmarkian adaptation theory, considered less trustworthy, as it was about validating the theory of Natural Selection itself. These are examples of the ways that SSK scholarship focuses on scientific controversy and the ways that human agency among scientists and social and cultural processes construct, craft and negotiate what becomes entrenched as knowledge.  13  Scholars of SSK argue that social processes and human agency come to bear on the dominance of one scientific claim over another. In SSK, the stabilization of competing claims involves enrolling allies (either professional or others) in support of the value of a particular claim much more than it involves the use of evidence (Latour 1987, 1994; Longino 1990, 2002). Claims that diverge from the pre-existing knowledge structure will have greater difficulty enrolling support. Whereas discourses and claims that can be used to extend existing knowledge structures (Latour 1987), especially where claims are valued and functional in other spheres of research and activities (Fujimura 1992), become more easily embedded as fact.  14  This research brings into question the notion of a vouchsafed truth only accessible to attentive scientists and seeks to identify the underlying internal processes and external influences that construct scientific knowledge. Controversies in particular reveal different possibilities for investigating the natural world that depend on social and cultural contexts and conceptual landscapes. The social constructivism utilized in SSK analyses suggests that some transcendental natural world can never be discovered or verified beyond the confines of the social (Yearley 1994). This body of work argues that it is impossible to entirely remove social processes from the idealized and rational practices of science. The claim made in However as Yearley (1994) notes, the study of controversy does not address the content and products of normal scientific activity that is taken as non-controversial and obvious. Fujimura (1992) illustrates the ways that scientific theories become stabilized through the entrepreneurial enrolment of non-scientific allies. She analyzes the production and dissemination of the oncogene theory. In its production she reveals how the entrepreneurial activities of two key scientists actively co-constructed the oncogene theory and standardized genetic material to be used in different cancer research programs and other areas of research globally. The ability to transmit a plausible theory with standardized materials that 'fit' within other arenas of research contributed to 'the oncogene bandwagon'. What this study reveals is that the validity and dominance of competing scientific claims is not only dependent on processes of validation in scientific circles but the value of a claim is simultaneously negotiated in other scientific and non-scientific areas of research and use. 13  14  15  SSK research is that the agency and negotiation among scientific actors in the production of scientific knowledge suggests that there is neither one correct path to access truth in nature nor that there is one correct truth in nature as argued in realist accounts but that instead 'facts' are socially constructed (Clarke and Fujimura 1992). For some, this brings material reality into question altogether.  15  2.3.1 Institutionalized Continuity in Science Latour (1993) moderates the understanding of social realism and the possibilities for the agency of scientists as the main or only force in shaping and stabilizing claims. In his book  We Have Never Been  Modern, Latour (1993) argues that neither the agency of actors nor the movement of facts can operate freely but that certain conceptual and practical institutions constrain and confine agency in science. For instance, categories of description and existing norms and discourses determine what data are considered appropriate and the value of how data 'fit' within existing conceptual categories and structures and contributes to the notion of progress in knowledge development. Similarly, Downey (1992) applies Goodman's argument that the continuity generated through scientific institutions is often underrepresented or downplayed in the social studies of science where social processes dominate empirical analysis. He asserts that this overemphasizes human agency and the authority that scientists have to influence epistemology and effect change. Instead he highlights the habit and continuity involved in the scientific institution; "Every new scientific theory, for example, reproduces a variety of assumptions, including theoretical and methodological categories and an understanding of theorybuilding as a legitimate, epistemologically significant enterprise" (Downey 1992, 73). This transforms the emphasis of studies of science from the interests of actors only, to include pre-existing conceptual schemes, assumptions and categories.  Other SSK research shows that there are self-referential and self-vindicating dimensions to closed system experimentation where "interests serve as standards against which the products of such extensions, new conceptual nets, can be assessed" (Pickering 1992). These self-referential and self-vindicating dimensions exist in disciplinary structures as well, where interests, norms and discourses provide the  The claim is that natural facts cannot be discovered through the rational natural sciences but are instead constructions of the social environment. This suggests that the interplay between the social environment and the construction of scientific facts needs to be studied by social scientists. Implicit in this argument is the idea that the hierarchy of the sciences needs to shift from the authority of the natural sciences to the authority of the social sciences. Highlighting the methods of the social sciences in this way instigated the 'science wars' in the 1980's. However this approach was brought under similar scrutiny with the contention that social scientists utilizing sociological principles to explain the work of natural scientists also fell into the same problems of observation, explanation and validation that were being challenged in the natural sciences (Ashmore 1989). This led to a conundrum where natural realism and social realism vied for positions of epistemological authority. 15  16  referencing structures in which new information must 'fit' (Downey 1992; Latour 1987). Theories, data and instruments are all part of this closed self-vindicating system whereby these elements dynamically interact and are calibrated until they combine in ways that make sense (Hacking 1992). These are some of the processes "where shaky and uncertain discoveries cumulate, gain authority and enter the sure path of science" (Latour 1987).  These analyses suggest that the dominance of one scientific claim over another is based in part on habits and institutions in science that contribute to its continuity and on the stability of pre-existing theoretical foundations. It is suggested that the idea of theory building in science replicates some assumptions (and values). This conceptual continuity therefore moderates the influence of the interests and agency of the scientists (Downey 1992). Both the questions asked in science and the methods used to validate theories are built from pre-existing conceptual schemes, assumptions and categories. Whatever theory or belief a practitioner of science holds influences and determines which facts are legitimate or 'fit' with the theory. This has been referred to as a self-referential system where theories, methods and data are simultaneously interacting. (This same understanding has been investigated and may be problematic in the modelling sciences where theories, data and instruments can become calibrated in a reinforcing and self validating system (Hacking 1992)). However the observation and reason of an individual is supplemented by social processes such as critical discursive interaction and norms that apply to practices of communities' not just individuals (Longino 1990). Therefore conceptual schemes and social practices of the scientific community also influence what is considered to be acceptable scientific work.  Whether there is "truth" or some natural world that exists beyond us is not a legitimate question to wrestle with here. Natural realists argue that the natural world can be discovered using rigorous and rational methods. At the polar end of SSK, social realists argue that there is no transcendental world but instead all categories, objects and phenomenon are social constructions of the mind. What is important to understand for the purposes here and what SSK reveals is that attempts at discovery and verification of natural reality are mediated through cognitive, social and professional processes. This undermines concepts of a value-neutral rationality and naive realism. However does this suggest that all forms of knowledge are equally valid? It is assumed here that, despite implicit social and cognitive processes, science is indispensable as it offers practices and a form of organization that provides standardized, systematic and analytical methods that describe and construct the world in useful and unique ways, usually with the aim of reducing it to a logically-consistent system of order. Scientific processes of peerreview, efforts at replication, and conscious attempts to minimize subjective views give rise to a uniquely reliable form of knowledge and that has had a considerable impact on framing and shaping our world. Can a research approach be derived that acknowledges the social, cultural and conceptual commitments in science while also acknowledging the effective though not entirely objective, rational or neutral  17  practices of science? Is it possible that something exists in between the two polarities of natural and social realism?  2.4 Science as Practice Recently, research into a new branch of SSK known as  science as practice has moved from the  understanding that scientific knowledge is based on the agency and cognitive commitments of scientific actors toward a more fluid understanding of science as craftwork and production. Researchers who examine the practice of science not only consider conceptual and social influences but look to the ways that the conceptual, social and material combine through ideas, discourses, instruments, materials, practices, institutions and agency to influence the outcomes of scientific inquiry. Science as practice is not bound to any disciplinary affinity but numerous disciplines are able to use constructivist and ethnographic approaches to analyze the practice of science rather than the theory of scientific knowledge. This approach is important in relaying the understanding that what scientists "do" is an important determinant of the knowledge they produce (Pickering 1992).  The interplay between subject, instrument and object are indistinguishable in science as practice. This approach combines with and advances the SSK literature and facilitates an investigation of how dynamic, non-linear and multi-scale systems are brought under the purview of science. 'Science as practice' highlights the laboratory and instrumentae as the border territory where the interpenetrating and interacting conceptual, social and material forces, generate 'facts' as we know them. Understanding science as "ongoing patterns of situated activity whose material is part of practice" (Rouse 1996, 14) offers a more realistic account of the interacting forces involved in and driving science. Using the laboratory as a unit of analysis, Knorr-Cetina (1992) probes the ways that the laboratory is a mere extension of the everyday world where social relations and 'tinkering' become critical for the mutual accommodation between actors, instruments and materiality. Knorr-Cetina (1992) asserts that these types of instrumental and technological transformations cause a disconnect between the objects and their natural environment (no regard for weather, season, location or viewer particularity). It makes thenimages continually present in a standardized format and in this way reduces stellar time scales to socialorder timescales. "The power of the laboratory (but also of course its restrictions)" she argues, "resides precisely in its enculturation of natural objects" (1992, 118).  16  How objects and phenomena become  studied or enculturated is not the result of discovery "out there" nor is it necessarily the result of social  In one of his studies, Callon introduces non-humans as actors and that influence the network-building in aquaculture research to show how scallops did not perform or standardize under the rules and conditions of scientific inquiry (see Callon and Latour 1992). 16  18  interests and conceptual processes, instead it may be the result of natural and social forces dynamically meeting and interpenetrating as hybrids.  2.4.1 Science as Social, Conceptual and Material Hybrids Callon and Latour (1992) argue that the practice of science is neither the project of natural realism nor social realism but is a socio-technical hybrid. In this fluid research approach, objects are not seen as products of nature but instead are viewed as a combination of material and human agency. Callon and Latour re-articulate a symmetry principle "not to alternate between natural realism and social realism but to obtain nature and society as twin results of another activity" (1992, 348). The conclusion that emerges from this analysis is that the natural and social worlds create and recreate one another in a complex and non-linear sequence of interactions and relations referred to as hybrids (Latour 1987). This is not equivalent to realism; instead materialism as observed, described and interpreted represents a mixture of conceptual/social/material interaction. In other words, a major part of the way materialism is both pursued and constructed (through scientific practice) depends on the underlying values, observations, selections and cognitive frameworks of the practitioner and the value of the findings in different communities and social contexts.  To analyze the practices that shape and define knowledge, brings the focus from some transcendental ability to achieve rational and universal meta-narratives about the world (either through natural or social realism), into a particularistic location where actors, ideas, instruments, access to resources, professional incentives, phenomenon and objects are interacting, dynamic and provisional (Pickering 1992; Clarke and Fujimura 1992; Nader 1996). This understanding of the practice of science thereby acknowledges social processes in science without privileging those processes over other factors that contribute to the way a claim is constructed. It identifies the particular contexts in which scientific information is constructed and the idiosyncracies particular to those contexts. This becomes important in understanding the assumptions, tools, "tinkering", behaviours and practices that contribute to claim construction (Clarke and Fujimura 1992; Pickering 1992). This understanding gets beyond the rhetorical or idealized descriptions of science as either natural realism or social realism and provides an alternate approach for examining the status of scientific knowledge. This approach has not been widely used however it becomes important for examining the role of science in society.  19  2.4.2 Science as Context-Dependent Often the most productive areas for enrolling allies in a particular claim reside outside the institution of science and in other areas of use. The value of what information 'fits' and 'does not fit' therefore becomes not only a project of the "core-set' but is also influenced by the values and'needs of broader networks such as funding agencies, research programs and agencies, and other fields that link to and use scientific data, studies, and/or assessments; "These networks are where shaky and uncertain discoveries cumulate, gain authority and enter the sure path of science" (Latour 1987, 103). The perception that science is inoculated from the social world is no longer a plausible rhetoric. Public and private sectors provide resources for scientists; selection committees for grants and sponsorships, and scientific actors simultaneously shape and respond to scientific trajectories. Yet scientific autonomy from the public sphere is considered necessary in order to ensure that fact creation remains objective, rational and neutral and effectively inoculated from external influences. Gieryn comments that the greatest threat "is the capture of science by policymaking powers - a loss of scientists' control over their research agendas and, in the limiting case over, what is represented as "scientific knowledge" (1999, 436). With the conventional 'policy for science' model that coordinates policies to fund new research programs, and the 'science for policy' model that uses science to shape new policy directives, it becomes difficult to know how much research drives society and how much society drives research; science and society interact in dynamic ways.  17  The discursive authority of science and its epistemological and rhetorical strategy provide the basis of its social power, and consequently, its institutional and professional privileges (Nowotny et al. 2001, 52). Once science becomes demystified we are able to understand that social processes, values, judgments, and ideologies all occur within the process of making and interpreting scientific claims. It is at this point that we are able to see a reinforcing social structure that emphasizes and possibly requires science to stabilize social order in a highly complex world where competing interests result in power politics. What science as practice research suggests is that inquiry is rooted in particular social, cultural and political contexts. The understanding of science as dependent on its social and cultural context thereby brings into question the universalization and hegemony of scientific knowledge. It has been argued that this universalization erodes the very real cognitive, cultural and contextual diversity that has led to the evolution, development and advancement of different cultures (Douglas and Wildavsky 1982; Rayner and Malone 1998).  18  For instance an individual's ontological status in the world and the context within which  Highlighting the depth to which science is contextualized within society, Fujimura (1989) remarks that the shareholders of biotech companies are present in the scientific laboratory at all times. Whenever science overcomes particularized constraints and constructs generalizable information, this information must then be relayed back to particular and local contexts in socially desirable, culturally acceptable and economically useful goods, products, services and knowledge (Nowotny et al. 2001; Jasanoff and Wynne 1998). 17  18  20  she/he lives ultimately influences the kinds of questions that are asked and the types of technologies and practices that are instituted (Douglas and Wildavsky 1992; Dove 1996). The ontological implications of a universal, apolitical science denies very real particularistic, contingent and democratic understandings.  2.5 Socio-Technical Hybrids Since social and cultural commitments underlie every stage of knowledge production, an increasingly relevant question is; who participates in deciding and interrogating the assumptions, values and commitments? It is proposed in this research that both knowledge-producers and the knowledge-users must participate in discursive interactions in order to make collective decisions over what is considered to be objective (in the form of impartiality rather than rationality) and sufficient information as it is applied to the political/public sphere. The symbolism of this proposal is that of an expanded core-set or "extended peer community" beyond the scientific community only (Functowicz and Ravetz 1992). If science is involved in self-referential categories whereby determinations of the value and relevance of information is not only left to the scientific community but is expanded, then 'knowledge users' have the opportunity to transparently interpret and negotiate general scientific findings back down to their particular environments. Longino (2002) highlights the importance of a defined community determining the adequacy of standards and methodologies rather than relinquishing it to some transcendental understanding. She asserts,  The point that there is nothing further, that appeal to standards and methodological norms beyond those ratified by the discursive interactions of an inquiring community is an appeal to transcendent principles that inevitably turn out to be local (Longino 2002, 174). In other words, an inquiring community defines what is epistemologically valid and determines which standards and methodologies are considered appropriate through discursive interactions and negotiations. Yet an appeal to the universal must also be aware of the local or context dependent character of knowledge creation and the ways that knowledge is both from and negotiated back into local contexts. In areas of policy relevant science, science that is used in the policy sphere to facilitate policy choices, the inquiring community, whether considered or desired or not, generally expands to include policymakers as knowledge users and in many cases interest groups also interested in determining what information is objective and adequate for the particular context or problem at hand. It is therefore reasonable to speculate that collective choice about the objectivity and adequacy of the information is defined both by the knowledge-producing community and the knowledge-user community.  21  In the book Re-Thinking Science, Nowotny et al. argue that, "under contemporary conditions the more strongly contextualized a scientific field or research domain is, the more socially robust is the knowledge it is likely to produce" (2001, 167). Social robustness is identified as relational whereby it depends on understanding the ideas, materials and interactions involved in the construction process (as described above). It describes a process, that in due course, may reach a certain stability. Nowotny et al. (2001) maintain that a distinction must be made between the robustness of the knowledge and its acceptability among different communities.  19  The "robustness is produced when research has been infiltrated and  improved by social knowledge... and is subject to frequent testing, feedback and improvement because it is open-ended" (Nowotny et al. 2001, 167). This understanding previews possibilities for more inclusive and transparent processes and scientific institutions that extend the peer community involved in delimiting and evaluating conditions that meet standards of objectivity.  2.6 Implications of the Analysis  If philosophers continue to believe that scientific knowledge is devoid of the social and sociologists focus only on relativistic understandings that place science as a mere construct, then the ability to discuss both the benefits  and the limits of science gets lost in debates over what information is considered to be  legitimate, correct and good. Although it has been argued that the identification of social processes in scientific knowledge and its methods undermines notions of rationality and objectivity, it does not undermine the effectiveness of science as a series of practices. Some scholars argue that exposing the tacit social, and thus subjective, processes inherent in scientific inquiry is necessary and can increase the reflexivity in scientific practice (see Ashmore 1989) and can highlight both the potential and limitations of science in a more transparent way especially with regard to public and political decision-making (Jasanoff 1996, 2003; Wynne 1995, 1996, 1998; Jasanoff and Wynne 1998; Lidskog and Sundqvist 2001). Using the momentum generated by the social studies of science literature, it is understood for the purposes here that both scientific knowledge and scientific practice are imbued with contextual factors including social and cultural commitments. These become erased or "purified" in the formalization of scientific products and are only identified through an examination of the process of science rather than its products. From the social studies of science understanding, scientific consensus and facts are ambiguous rather than absolute. The use of constructivist methods helps to differentiate between what is theorized  Civic science argues for broader non-scientific participation in knowledge production (Rayner and Malone 1998a). It, however, remains unclear to what extent broader stakeholder communities can be integrated into inclusionary processes while maintaining efficiency and effectiveness of decision-making 19  22  about in scientific knowledge and the ways social, conceptual and material processes and practices craft it  in situ. 2.7 The Following Chapter  In this chapter it has been argued that research into the practice of science argues that social and cultural assumptions and processes underlie all stages of knowledge production and therefore inscribe normative commitments into the framing, interpretation and representation of scientific information. Social, conceptual and material variables intermix in the process of claim construction. This becomes important for the following chapter where the role and use of science in society and in particular, policymaking is examined. If the non-mandated sciences are the product of hybrid practices that combine rational and social processes that depend on the practitioner and the social and cultural context, the suggestion can be made that where scientific representations and products are used in the policy sphere, these tacit assumptions, commitments and meanings are also transferred into the socio-political sphere. If science and scientists cannot provide entirely objective, rational and neutral descriptions of the natural world, what is the role of science in the policy sphere? This question is explored in Chapter 3 where the conventional use of science in policy is brought into question. Conflicts that arise in technical decisionmaking between scientists may not be the result of disciplinary differences but may in fact be due to the ways that different data, observations and 'facts' can be interpreted in different ways. In public controversies, the flexibility of science supporting opposing sides of the controversy enhances rather than reduces decision-making complexity. In the following chapter a normative and pragmatic framework is proposed to develop a systematic process that attempts to account for these normative commitments in the production and use of scientific information.  23  C H A P T E R III: Bridging the G a p : Problematizing International Science  Examples from environmental science research reveal how the obstruction or erosion of jurisdictional boundaries between science and policy can lead to a loss of legitimacy in either community (Nelkin 1992; Sarewitz 2000; Pielke 2004). This is often considered to be the difference between a technocratic or scientized politics or a politics of science (Elzinga 1997; Jasanoff 1990, 1996b). Scientists often charge the policy community with the misuse of and piecemeal selection of scientific information to suit preestablished policy goals. In contrast, policymakers often accuse science of not being relevant to the needs of policymakers in efforts to respond to current issues.  A n interesting side-effect of this battle over information and its requirements is that, although these disputes are not resolved, it usually results in increased budgets for science, as calls for more research are supported by policymakers in their search for relevant and robust scientific information and are supported by scientists in their efforts to secure funding for future research (Sarewitz and Pielke 2000; BoehmerChristiansen 1994a,b). This is particularly true of environmental problems that are typically accompanied by high scientific uncertainty due to the multi-causal, non-linear and dynamic aspects that make it difficult for the tractability of environmental science.  20  That science and policy communities mix ideas  and beliefs with values, norms and institutions should not blind us to the differences between social practice and organization between the two domains (Miller 2001). "A culture or moral economy of credibility may operate in each context, but we would not expect the cultures of the laboratory and the legislature to necessarily exhibit the same characteristics" (Miller 2001, 482). The view that science can directly contribute to the resolution of societal problems, although problematic in various ways, cannot be dismissed. Science, with its unique processes and practices, are required to address increasingly complex and multi-layered technical and environmental problems. Yet this response alone is inadequate. Further research into pressing public policy problems at local, regional and global scales will only be helpful if the scientific understanding it produces can be understood in its contextual richness and connected in some useful and interactive ways to the policy process.  This chapter will briefly explore a prominent understanding of rationality in the socio-political sphere and the failures of the scientization of policy. Using the constructivist understandings of the previous chapter, this thesis proposes a framework that operationalizes the view that in areas of mandated science where science is selected, aggregated and used in the governmental sphere (Salter 1988), hybrid discourses are  For instance determining what level ofriskis adequate for genetically modified organisms becomes problematic once strains move from the confines and controls of the laboratory to the market and into open and dynamic systems.  2 0  24  produced that are neither science nor policy. It will be argued that traditional efforts to 'purify' the domains between science and policy may not only be impossible but may also be undesirable. Instead, it is proposed that an approach that considers the flexibility and interpenetrating relations between these communities can more adequately provide insight into the socio-technical nature of increasingly technical public policy problems (i.e. health care provision and environmental problems) than attempts to maintain science and policy as. discrete and separate realms. First an analysis of the institutionalized arrangements between science and policy at a domestic scale is explored in order to identify the tasks, responsibilities and expectations delegated to each community. Problems associated with the philosophical characterizations of science in policy relevant settings are investigated to establish a rationale for introducing constructivist methods and understandings. The conventional scientization of policy, where science is used to legitimate policy decisions, often leads to the balkanization of scientific interpretations (Sarewitz 2000). A corresponding politicization of science therefore becomes a secondary effect (Pielke Jr. 2002). Second, areas of policy relevant or mandated science are explored as hybrid areas where the transfer of scientific information to the policy sphere fundamentally changes and alters the nature of the science itself. Mandated science reveals the ways that the discourse in the policy arena is neither scientific nor policy, it is a hybrid of the two. Third, the notion of co-production is introduced to suggest that despite attempts to purify the roles of science and policy into distinct boundaries, science and policy are interpenetrating social worlds that simultaneously coproduce natural and social order (Jasanoff and Wynne 1998).  A theoretical framework that attends to the interpenetration and interdependence between science and policy while simultaneously maintaining sources of credibility and legitimacy proves to be a paradoxical challenge. Possibilities for deriving science assessment processes that include knowledge producers and policy users in the production of policy relevant scientific information will be proposed in the Imbued Meaning framework introduced in this chapter. This theoretical framework utilizes lessons from the social studies of science and from the failures of the scientization of policy in national contexts to outline increasingly pragmatic ways of conceptualizing credibility, legitimacy and relevance in global environmental regimes.  3.1 Models of Science in Policy The taken-for-granted nature of science has been brought under a critical gaze for decades in US politics. A critical discussion of the failures associated with the scientization of policy in the domestic sphere should begin with a discussion of the underlying reasons why science was integrated into the policy25  making process and why scientists and scientific information have developed such a significant presence in the socio-political sphere.  3.1.1 Science for Policy  In 1946, Vannevar Bush, US Director of scientific research and development, submitted a report entitled "Science: The Endless Frontier" to President Truman (Bush 1946). This report called for a post-war program for scientific research. Through consultation with White House committees relating to health, public welfare, the development of scientific talent and publication of scientific information, a strong case for the value of science in the public realm was made, highlighting the untapped benefits that science could develop on behalf of society. In this report, Bush documents the endless combination of human discovery and human ingenuity through science and technology and the ways it could contribute to the continual advancement of society. It was assumed that by understanding the natural world through science and harnessing its energy and creativity through technology humans could achieve greater technical, political and social advancement. This was viewed as a way to provide greater control over the unpredictabilities that were posed by both the biophysical and political world (ie. disease, threats of natural disaster, national security, public health, etc.). He convincingly argued that these considerations meant that greater national spending should be directed to science and its possibilities for the sociopolitical sphere. Not only could science directly aid American society by advancing knowledge and technology but also according to Bush, based on objective and value-neutral scientific findings, decisionmakers could access a greater reservoir of information, a resource base to pull from in order to make better policy decisions (Bush 1946). Through the development of a techno-scientific agenda, it was assumed that science would endlessly supply the resources required to make human life more comfortable and convenient (Jasanoff and Wynne 1998; Shove 2003; Bush 1946), thus creating a greater cultural presence for science in society. With this increased cultural presence for the scientific enterprise  21  .  (Gieryn 1983, 1999), rational and technical knowledge became the rubric of democratic societies and decision-making practices.  22  Science became a political resource, where the rational, autonomous and  value-neutral information it was thought to produce was used to legitimate and justify policy decisions (Price 1965, Baumgartner and Midttun 1987; Robinson 1992, 1992b).  This is part of a longer history of the extension and mobilization of science as a profession and the ways rhetorical justifications such as objective, value-neutral, rational, relevant, etc. have been used to ensure the continued cultural presence of science within society (see Gieryn 1983, 1999). The connection of science to policy was initially conceived as a socialist agenda in the Soviet Union in the 1930's. Cognitive authority and power shifted from that of elitist politics to the objective, rational and neutral methods of science (Jamieson and Elzinga 1997). 2 1  2 2  26  3.1.2 Truth Speaks to Power  The role that science has played in society has been influential and dominant in social and institutional decision-making structures (Gieryn 1999; Jasanoff 1990, 1991; Jasanoff and Wynne 1998), due to the prevalence of what might be called a 'truth speaks to power' (Price 1965) view of the science-policy relationship. The 'truth speaks to power' model of science policy is based on a similar view to Bush (1946) whereby science was viewed as a way to access truth in the natural world that could then be used to legitimate policy decisions among powerful officials. The perceived apolitical status of science has resulted in a great cultural presence within society (Gieryn 1999). The precise, reductionist and rational quality associated with the products of science is seen to provide information that is useful for the policy sphere, because is it accurate and is not tainted with political interests or biases. Policy analysts use scientific research and studies to determine the causes, effects, mitigation and remediation associated with public environmental policy decisions (Jasanoff 1986, 1990; Pielke 2000; Miller 2001). Analysts often view scientific information as objective, discrete, and value-neutral. As such, it becomes an indispensable political resource that can provide important information, and also the legitimacy required for the fuzzy arena of public policy and decision-making. Public issues regarded as controversial or problematic in politics are often put into what is perceived as the objective, value neutral and rational hands of scientists and scientific inquiry.  The phrase "truth speaks to power" was coined by Price (1965) to indicate the unidirectional flow of information or "truth" from the autonomous scientific community through to the political or "power" communities. This model of science and policy views these communities as separate, distinct and distinguishable, each operating within their own sphere of authority and practice. Science uses routinized methods, autonomous from the political sphere to decode the universe around us and to pursue truth. In contrast, politics is comprised of interests, values and power relations to determine the 'best' course of action on behalf of its respective constituents.  23  Price (1965) addressed the problem for science and  policy as that of the translation of science into policy. In other words, he examined ways to retain the integrity of the original scientific research as it moves into areas of informed policy making. Price's account provides an idealized understanding of the ways the various professional, administrative and political estates define and constitute their professional norms to translate scientific information (Jasanoff and Wynne 1998). For instance, he suggests that the political estate has an interest in preserving the integrity of science due to a political code of conduct that promotes the view that policymakers must remain responsible and accountable to their public (Price 1965).  These descriptions purposefully use the terms science and policy as nouns to highlight the unitary and distinct ways these communities have been conceptualized. 2 3  27  However, this idealized model has increasingly come under pressure. Over the last several decades, the attitude toward science and policy has shifted from this steadfast faith in democratic institutions and essentialist constructions of knowledge. For instance in his account of nuclear missile guidance systems, MacKenzie (1990) introduces the idea of a 'certainty trough'. Policy use and dependence on scientific information tends to bury associated uncertainties and assumptions related to original research and the original research context. He argues that once science moves from its original scientific context and goals, further from the original practitioner and closer to the policy sphere scientific assumptions, approximations and uncertainties are minimized and become increasingly idealized. Without attention paid to the provisionality of science and its qualifications, policymakers can haphazardly (or malevolently) 'purify' scientific information and use it to authoritatively legitimize and serve a particular, often predetermined policy. McKenzie (1990) also highlights the ways that an increasingly politicized and sceptical civil society has placed new and at times contradictory demands on democratic institutions. Increasing political pressure has exposed the flexibility internal to science. The diverse demands of science in the political sphere have exposed the ways that science can be used as justification or legitimation for opposing political arguments (Wynne 1982, 1992a; Robinson 1982a, 1992; Baumgartner and Midtun 1987; Salter 1988; Jasanoff 1991). It is not uncommon to see scientists arguing on opposing sides of a policy debate thus placing additional pressures on scientists operating in the policy sphere.  Even in situations where consensus-based science is provided to inform the policy sphere, this information can often be overridden by other political and social concerns.  24  If truth does speak to  power, than it might be expected that consensually agreed to scientific information would regularly be applied to policy decisions however this is not necessarily the case where counter experts can be mobilized to interrogate scientific consensus, ultimately contributing to a catch-22. The 'truth' is desired to reduce conflict and to legitimize decisions and yet by virtue of truth being transferred to areas of conflict and decision-making it no longer retains the appearance of truth. This challenges the unidirectional 'truth speaks to power' flow of information.  A recent example if this is the Canadian Species At Risk Act (SARA) (www.speciesatrisk.gc.ca/default_e.cfm) which ensures that a panel of biological scientists, the Committee on Status of Endangered Wildlife in Canada (COSEWIC) determines which species are at risk, threatened, and endangered in Canada. However the SARA legislation also ensures that discretionary decisions of the Cabinet determine which of the 184 endangered species identified will receive federal protection based on political estimations of costs and benefits (www. sararegistry. gc. ca/the_act/default_e. cfm). 2 4  28  3.1.3 Politics of Science  The result of the use of science in the policy sphere is that it is common in the policy relevant environmental sciences to find controversies over the sufficiency and adequacy of information and its requirements. Where science is presented as unified and robust, counter-experts can be mobilized to highlight uncertainties and assumptions in the science. Counter experts can either highlight uncertainty and the need for more robust research or they can mobilize contrary evidence using information from the other side of an unsettled scientific controversy or from different disciplinary vantage point (Salter 1988; Brunk et al. 1991; Jasanoff 1990, 1995, 2003). When the 'battle of the experts' begins over the use of science in the socio-political sphere it causes confusion over legitimate courses of action. These controversies often end in deadlock, and when a decision is imminent the authority of the policy decisions becomes embroiled in a power politics (Sarewitz and Pielke 2000). Robinson (1982) goes further to suggest that in the public domain, ethical and normative arguments are often framed as scientific and technical questions. Any values that underlie a particular policy can be represented in technical terms with questions of value framed as questions of fact. In this way scientific information can be used to support a variety of ideological and value perspectives both within the policy and the scientific sphere.  This flexibility and plurality in scientific evidence available for the policy sphere has contributed to a relativization of science  25  in the adversarial politics of the US (Jasanoff 2003). Where scientific evidence  cannot authorize a unitary decision, policy reverts back to the arena of power politics. Underlying the mis-use and relativization of science are fears that scientific advice about environmental matters may be too costly and/or may reduce global competitiveness and advantage (Harrison 1996). Lobby groups and the power differentials among them compete for the attentions of the administration using evidential scientific selections to justify their positions. This struggle for resolution can become strategically relativistic where the interests of a political administration override scientific evidence by claiming that it is insufficient or inadequate or by hijacking the scientific process altogether. For instance, the Republican Administration under George W. Bush has been charged with 'fixing' scientific panels that ensure politically authorized scientists comply with policy directives.  26  The point is that attempts to  scientize environmental policy issues, especially in the adversarial environment of US politics, paradoxically fuels the politicization of science.  The relativization of science has negative connotations. This is understood as a political or social belief that science cannot contribute understanding to the highly complex, non-linear and dynamic aspects of public policy problems. For instance in Health and Human Services (HHS) (Jasanoff 2003) and in the National Academy of Sciences Climate Report (Agres 2004; Union of Concerned Scientists 2004), allegations have been issued against the Bush Administration claiming 'fixed' panels in the former and misuse of experts in the latter. 2 5  2 6  29  The politicization of science and scientization of politics are thus two faces of the same coin. Political rationality puts science into the precarious situation of being labelled either 'good' or 'bad', objective or advocacy-based, robust or uncertain. Policy officials downplay underlying uncertainties and assumptions when and if scientific information supports their particular policy decisions. In contrast, where information does not support particular policy goals officials or interest groups can highlight the uncertain areas inherent in scientific research in order to delay policy action. What this exposes is that cognitive authority does not entirely reside in the arena of science. Instead the legitimacy and value of the science is a negotiation that takes place at the science-policy interface (Jasanoff 1990, 1996).  3.2 Policy Relevant or Mandated Science  The perceived advantage of science operating on behalf of the public sphere is that agreement among scientists has the effect of removing areas of discourse from ethical and political contestation (Jasanoff 1996a, 174). The origin of this use of science in policy is linked to a political agenda that attempted to ensure that adequate distance existed between policy and politics (Jamieson and Elzinga 1997). In this perspective, the objectivity and accuracy of scientific methods used to calibrate the truth existing in the natural world are also used in the policy sphere to inform rational policies in the social sphere. The disadvantage of this scientization of politics is the way that it is assumed that scientific information is inoculated from ethical, moral and democratic understandings despite the implicit values in science have on the socio-political sphere. Liftin highlights the paradox of the scientization of policy," [s]ince science is modernity's pre-eminent instrument of legitimation, all participants can be expected to claim that their positions are mandated by science, even if science alone can never mandate anything" (1994, 6).  3.2.1 Problems with Translation  In response to these kinds of considerations, Weinberg (1972) located an area between science and policy he called "trans-science". This was an area where policy questions could be asked of science but not answered by science. His concern was that scientific information could not be transposed from its original context and purpose while being directly relevant to the needs of the policy community. Instead he recognized the ways that information had to be extrapolated, managed and moved within this hybrid area of trans-science. As Jasanoff and Wynne note "Weinberg's essay can be seen as a defining historical moment when the hybrid character of science done for policy became briefly apparent within the US  30  scientific community" (1998, 9).  Science used for policymaking has more recently been referred to as  an arena of 'mandated science' (Salter 1988). Mandated science can be identified when a scientific panel or advisory body has, as its primary audience, a governmental or regulatory body that seeks recommendations from it (Salter 1988). Scientific work, either in the form of individual studies, scientific testimonials or scientific assessment, becomes "mandated" when it is evaluated in terms of the conclusions it can offer to policy makers about the merit of particular regulations (Salter 1988).  Using four empirical case studies on standard-setting, Salter (1988) identifies four defining characteristics of mandated science. First this type of science has a governmental audience waiting for recommendations from it whether in the form of an individual scientific study, a testimonial, an advisory body or an assessment forum. Second, despite the connection to the socio-political sphere, science, scientists and scientific evidence tend to be idealized when used either in the political and/or legal domains. Third, mandated science straddles the boundaries of the scientific and policy worlds and therefore embodies a unique form of discourse that is neither solely scientific nor solely political. The fourth characteristic is described as the moral dilemma in mandated science, attributed to the extension of values and particular cultural and moral commitments of scientists as they extrapolate studies and research into the broader socio-political domain (Salter 1988). Similar to Weinberg, Salter views this as a different form of scientific activity than that associated with 'normal' or 'pure' science. In mandated science, scientists must align themselves with the policy community and grapple with the fact that science "is directed toward closure, towards the production of conclusions that would support decisions taken in another sphere of activity, government or economic relations" (Salter 1988, 189). Mandated science is a unique area where scientific expertise has direct implications on normative questions of fairness, democratic rights and moral issues in the consideration of scientific outcomes.  Whether science for policy is pure or applied, or whether it is in the form of selection and synthesis for a political audience, scientists are put under a unique kind of pressure to translate scientific studies clearly, simplistically and concisely and in order to "facilitate clear choices" for policy (Salter 1988, 5). In the situation where information on a particular topic is to be integrated into a single scientific report, scientific interpretations are transposed from a complex, contested and uncertain setting to a setting that demands concise, simple and consensual knowledge. Due to the burden of proof, the scientific canon of hypothesis testing rests on the doctrine of "theory until proven fact" as a fundamental component of the truth-seeking process. In scientific disputes "a fundamental dichotomy (exists) between the potential dispute resolution objectives of 'truth' and 'justice'" (Salter 1988). This method makes it difficult to  Weinberg advocated making a distinction between 'pure' and 'applied' scientific practice in order to retain the autonomy and credibility of the scientific activity ('pure' science) while at the same time producing knowledge that is directly linked to policy problems ('applied science') (Jasanoff and Wynne 1998).  2 7  31  connect science to the often strongly normative decisions required in a policy context. The moral dilemma identified in mandated science is that experts, based on their scientific credentials, are asked to inform particular policy processes precisely due to their idealized insulation from the socio-political sphere, yet they must concurrently assume the role of advising on regulatory or social issues beyond their areas of expertise (Salter 1988; Jasanoff 1990). This dilemma highlights the problem of translation of scientific information to the policy sphere under the unidirectional model of science-policy.  By selecting and framing information for its policy relevance, scientists implicitly transfer underlying commitments and values and present information as though it is objective and value-neutral. To provide a hypothetical example, climate scientists know that banning CO2 emissions could result in extreme economic costs, but they also know that not banning emissions will increase the rate and magnitude of climate change. If asked to advise on such an issue, underlying ideological, political and moral judgments are implicitly made about the appropriate response. This is the paradox of science operating in the policy sphere; to be considered good science, scientific information must be constructed and presented without regard for its social or political implications. Yet, for the scientist to provide convincing evidence to both political and regulatory bodies, they must anticipate the policy situation and legal constraints under which it will be interpreted.  28  Scientists that choose to participate in mandated science are often looked upon suspiciously by scientific peers and the public as advocating a particular interpretation or agenda (Campbell 1987; Salter 1988; Jasanoff 1990, 1995; Boehmer-Christiansen 1994a,b). In the 'truth speaks to power' model, unambiguous science is needed to legitimate policy decisions. As noted above, when scientists provide their interpretations, counter experts can be mobilized to expose areas of extrapolation or uncertainty in the scientific interpretation for the social context. This leads to contestation and controversy, confusion over which information is correct, making policy decisions even more complex. Policymakers are forced to make decisions about which experts have authority rather than decisions about the issue at hand.  In public policy issues, different interest groups can use incontestable scientific evidence to argue from opposing sides of an issue and are able to use varied disciplinary affiliations to justify their arguments and claims (Yearley 1996; Social Learning Project 2001a,b). This problem is often attributed to different disciplinary goals, methods and objectives. For instance, an economist will examine a problem in different ways than an ecologist who will examine problems in different ways than a chemist, and so on. An example of this in the legal context is illustrated in Salter's (1988) example of the health and ecological consequences of lead particulate in the municipality of Toronto in Ontario (The Toronto Lead Controversy Chapter Five). The inability to derive "objective" or sufficient evidence pointing to the Toronto Refiners and Smelters Company in the community as the absolute cause of the documented health outcomes meant the judge ruled in favour of the defendant and called for more 'objective' research.  2 8  32  In principle, disciplinary differences in methods and goals can be dealt with through issues of cooperation and translation. However in practice, (as discussed in Chapter 2) these differences are the result of the 29  deeply embedded assumptions, judgments and commitments of the practitioner and the social and cultural context in which scientific information is constructed. Once science is transferred to the policy sphere, these assumptions are also transferred to and negotiated in the policy sphere.  3.2.2 Value Frameworks in Mandated Science  Often controversies are framed as scientific where each side is armed with incontestable science legitimating two interpretations of an argument. The issue in these situations is less about the accuracy 30  of the science and more about the frameworks used to select and interpret the science. In other words, the emphasis becomes less about the content of science and more about the processes in which interpretations are derived. In situations of mandated science, what are left to the scientist, expert or assessment body are judgments about which science, which methods and which interpretations are appropriate for assessing and responding to a given problem. In this way the choice of which methods are appropriate and interpretations are acceptable are based on the normative judgments and value-frameworks of the assessor (Robinson 1982, 1985; Brickman et al. 1990).  Robinson's (1982) investigation of'hard' (business as usual) and 'soft' (alternative energy) forecasts about future supply and demand of energy demonstrated each argument was equipped with incontrovertible evidence about the merits their approach. Therefore, decisions regarding their approach were less about getting the 'facts' right and more about the underlying commitments and valueframeworks that are used to the select and interpret. Robinson summarizes the issue succinctly:  With respect to the specific calculations used, the important point is usually not the accuracy of those calculations but their interpretation and relevance. In turn, the interpretation and relevance also depend upon the premises subscribed to, literature read, and methods used (1982b, 32). What information is selected and how it is interpreted therefore lies with the prior commitments and value frameworks of the analyst. Robinson (1982) argues that what is perceived to be a scientific process, is often based on prevailing patterns of belief and assumption about the world and how it operates. He posits that the interpretation and use of data is based on a, "process [that] works at the level of premises not of "factual" data and it thus tends to be invisible and not subject to falsification" (Robinson 1982, 32).  Integrated assessment is an example of multi-disciplinary cooperation (Rotmans and Dowlatabadi 1998) For instance, Campbell (1985) analyzes the scientists and scientific evidence used to bolster both the proponents of the development of the McKenzie Pipeline and opponents to it.  2 9  3 0  33  In another study, Robinson and Hooker (1987, 132) find that the Canadian National Energy Board's use of forecasts became "both cause and consequence of policy decisions". They argue that forecasts base present decisions on the prediction of future events using current assumptions to determine the future and do not consider possible changes in behaviour or more varied supply of energy sources. Without considering the role of present choice, these forecasts give the appearance of inevitability to future energy paths and lessen the potential for creative policy options to present problems. The authors also argue that the use of forecasts allows decision-making institutions to assume "a false cloak of objectivity" (Robinson and Hooker 1987). In this way, when presenting scientific information scientists and experts were simultaneously presenting assumptions about the nature of social order and about the needs and values of the policy community.  What is recognized in this analysis is that premises can be based on a significant degree of expert knowledge and yet they are not unambiguous scientific facts. Robinson (1982) finds in his analysis of hard and soft energy forecasts that the premises or choices that 'fit' with existing beliefs, values, and commitments have greater authority in the policy sphere. He finds that 'soft' energy paths, that considered future variables and possibilities for decreasing demand through conservation measures and increasing supply via alternative energy, was seen to be unknown and uncertain territory and therefore less scientific. On the other hand 'hard' energy predictions that used familiar econometric (i.e. price and income) and quantitative approaches 'fit' within the existing value framework of the policy community despite being consistently inaccurate about future energy supply and demand (due to unforeseen influences).  31  The latter used contemporary supply and demand patterns to project into the future, thus  embedding present assumptions, data and value frameworks in future projections that reinforce current understandings about future modalities and limit options and possibilities.  Brunk, Haworth and Lee (1991) present similar propositions in their analysis of risk assessments and in particular the deregistration of the herbicide Alachlor in Canada (see Box 3.1 at end of the section). They find that the way risk is assessed and interpreted is influenced by how normative questions of what is fair, equitable and reasonable are answered.  In short, risk assessors who strive in the face of mild uncertainty to be strictly neutral and to attend only to the facts, cannot attain the neutrality to which they aspire. Whether they recognize it or not, any decision they reach will depend for its reasonableness on the relative worth of the various values that their ostensibly valuefree approach puts at risk (Brunk et al. 1991, 37). In this case supporters of the hard energy forecasts had enumerations, quantified variables (i.e. price and income) and what was considered to be objective data. This was one of the things that made this approach more valued than more qualitative and alternative approaches. These were considered to be the ambiguous and uncertain. 31  34  That normative issues pervade what is perceived as the quantitative, objective and value-neutral territory of scientific estimations of risk becomes significant. It undermines the intended role of science in policy - to present objective and value-neutral information that will lead to better policy decisions. Where nonscientific and non-technical judgments are tacitly made scientific risk assessment oversteps the boundaries of the technical into the territory of normative judgment or premises (considered to be the territory of risk managers). Underlying assumptions of the risk assessor prescribes specific conditions or make implicit assumptions about what is considered to be fair, equitable, and reasonable having implications for the policy sphere.  What this research suggests is that mandated science is a different type of scientific activity (Salter 1988). It suggests that experts supporting polar sides of an issue are arguing less about the scientific and technical information and more about the underlying ideologies, assumptions and extrapolations used to frame the problem. In other words, the controversies stem from the normative accounts of what technical information is considered sufficient, adequate and appropriate for the non-scientific sphere thereby leaving the underlying premises involved in the selection of information open to negotiation. As noted in Chapter 2, social and cultural commitments underline every stage of knowledge production not just in its transfer. Science and its underlying commitments carry particular normative values that may or may not be relevant, valuable or meaningful to a policy problem and that certainly become contestable in a public or policy environment. Where scientists, advisory bodies and assessments are shown to have underlying assumptions that are not about unambiguous facts but are about particular value frameworks, the unidirectional model of science for policy is brought into question. Whether to agree with or reject these assumptions becomes the project of the policy sphere. For instance in Brunk et al. (1991) analysis of Alachlor (Box 3.1 below), the government Health Protection Board (HPB) bypassed the advisory Review Board's recommendation to reinstate Alachlor. It therefore bypassed or rejected the underlying values associated with that recommendation in favour of a 'safety' or no-risk framework.  Despite idealized commitments to objectivity, mandated science is imbued with meanings and prescriptions that are extrapolated beyond the underlying goals and values of science (Salter 1988; Jasanoff and Wynne 1998). The underlying values are negotiated for how they 'fit' in the policy environment. Therefore the question becomes how does one utilize the analytical and descriptive qualities of science while being able to interrogate and assess the normative or prescriptive commitments of science in the policy sphere? This speaks directly to the use of certifiable and trusted scientific experts in the public policy sphere and the processes in which information is selected and interpreted for the policymakers.  35  Box 3.1 The Multivalency of Science in Policy: The Case ofAlachlor in Canada Brunk, Haworth and Lee (1991) in their study of the risk assessment on Alachlor, an herbicide used on soybean and corn crops, found that the same scientific information, epidemiological and skin absorption studies, were interpreted differently by the parties involved. They found that these differing conclusions were not due to the acceptance or rejection of scientific information; all were using the same scientific information, nor was it due to the differentiation of'good' or 'bad' science. Instead it was due to the different value frameworks involved in interpreting the information. The risks Alachlor posed as a carcinogen were interpreted and disputed by the three key parties: the federal government Health Protection Board (HPB), Monsanto Corporation and a non-partisan advisory Review Board charged with making objective recommendations. While the HPB interpreted Alachlor as posing an unnecessary human health risk and "safety" issue due to the small (and statistically insignificant) number of rats found with tumours at low doses, the Monsanto Corporation suggested that the human exposure to Alachlor was low (approximated at five days a year) and therefore the herbicide posed no significant risk if exposure was amortized over a human lifetime. This dispute over tolerable risk revealed that the HPB's interpretation was based on the view that no risks were acceptable (a zero-risk criterion), while Monsanto argued that the risks were low based on minimal exposure to the herbicide. Monsanto directed the argument from safety to economics arguing this was an unfair penalization of Alachlor and that other comparable herbicides on the market such as Metalochlor, produced by a rival company, should be included in the review. Monsanto argued that Alachlor and Metalochlor posed similar risks and yet the deregulation of both would have significant economic consequences and pose risk for economies of scale and global competition for Canada. They posited that the most reasonable way of determining risk in this situation was through the use of riskbenefit analysis (what level of risk is tolerable in order to maximize economic benefits). Weighing the information and these arguments, the Review Board agreed that Alachlor posed a risk but believed the government over-estimated the risks and sided with Monsanto that a risk-benefit analysis should be used. This conclusion thwarted the HPB's attempts to make an assessment purely on safety issues shifting the focus to the economic costs and benefits of Alachlor.  Meanwhile other interest groups such as a civil lawyer representing a farm-wife, Mrs. Zelma Van Engelen,fromLambton County, Ontario and environmental groups were concerned with the detrimental effects of chemical herbicides on human health and the ecology of farm regions. Mrs. Van Engelen was a breast-feeding mother who became concerned about Alachlor when Ontario Ministry of Environment officials had discovered the herbicide in her well at levels considered dangerous by the federal government. In the ruling of the Review Board to use a risk benefitframework,understandings of zero risk were undermined. The burden of proof was shifted to the risk cautious (the HPB, Engelen, and environmental interests) to justify its view of the unacceptability of risk. In thisframeworkthe uncertainties about the safety of Alachlor worked against the risk-averse and favoured Monsanto. Underlying the Review Board's claims to be rational, neutral and objective assessors of the facts were three normative assumptions about the world. These assumptions related to 1) the nature of social order, 2) the value of technology, and 3) understandings of rationality. What Brunk et al. (1991) found was that the Review Board and Monsanto had similar underlying value-frameworks. This framework related to assumptions based on 1) a liberal politics focussed on the notion of progress and risk taking; 2) a pro-technology stance which viewed the scientific product of Alachlor as a benefit that needed to be used and advised that well liner technology be used by Mrs. Van Engelen to protect her well water from the run-offfromsurrounding agricultural lands; and 3) a focus on quantitative or "factual" accounts considered to be rational led to the marginalization of "normative sentiments" relating to human health and environmental values about the overall use of herbicides. They find in their case study is "it appears more is at stake in risk debates than can be settled with science alone" (Brunk, Haworth and Lee 1991,23). Despite the Review Board's recommendation in this case, the Health Protection Board decided in favour of zero risk and prevented the reinstatement of Alachlor.  36  3.3 Scientific Process versus Scientific Content  Stabilizing scientific information in the policy arena becomes less about the credibility of the content (each side may have incontestable evidence) but instead becomes more about who has the trust and credibility to select and interpret scientific information. Wynne (1996) questions the dualism between science and policy. He finds that analyzing 'sides' is a simplistic approach to complex problems where there are multivalent interpretations of any one particular context. In his seminal case study, the Windscale-Sellafield inquiry, he finds that the legitimacy of the information was based less on realist concerns about bio-physical processes and shifted more to social realist concerns based on the trust of experts and authorities in the inquiry process (Wynne 1989). Wynne comments;  Abstract scientific knowledge may seem universal, but in the real world, it is always integrated with supplementary assumptions that render it culture-bound and parochial. The validity of this supplementary knowledge crucially affects the overall credibility of 'science' or 'experts' (1989b, 12). i Similar to Robinson (1982) and Salter (1988), Wynne (1989b) argues that scientific content is less of a determinant in the decision-making process. Instead its validity is assessed through a parochial lens where issues of the competency of officials and experts, adequate decision-making processes and transparency become criteria that determine the validity of scientific claims. These types of social factors are viewed by a number of scholars to affect the credibility and legitimacy of global environmental assessment processes (Wynne, Simmons et al. 1997; Clark, McCarthy et al. 1997; Farrell et al. 2001; Clark and Cash 2002).  32  Process design has been highlighted as a way to increase the legitimacy of  scientific information through such elements of participation, fair representation of actors, and iterative assessment to name a few (Clark, McCarthy et al. 1997; Farrell et al. 2001).  In conjunction with the validity of experts, officials and the processes are also processes of science-policy interaction that occur at the interface where underlying value frameworks and supplementary assumptions are interrogated and the boundaries between science and policy are negotiated. In her empirical research on standard setting for pollutants, Jasanoff (1990) finds that in the arena of mandated science the domains of scientists and policymakers are not predetermined, but rather their respective boundaries are negotiated in situ at the science-policy interface. These negotiations are agreements about the role and authority of each community over particular concepts at the interface (Jasanoff 1990). Empirical studies that examine  Global environmental assessments are the entire social process by which expert knowledge related to a policy problem is organized, evaluated, integrated and presented in documents and otherwise to inform decision-making (Farrell et al. 2001, 312). In this definition assessments are viewed as areas of mandated science that embody hybrid forms of activity and that require credible and legitimate processes for transferring relevant science to the policy arena. 3 2  37  the role and use of science at the policy interface suggest that any effort to analyze policy requires the simultaneous analysis of science and vice versa (Jasanoff 1990, 1991, 1996, 2004; Wynne 1996, 1992; Jasanoff and Wynne 1998; Lidskog and Sundquist 2003). Policymakers and risk managers negotiate the value of the science for the particular and local context in which it is being used (Jasanoff 1990). In other words, scientific information does not float freely but instead determinations of what is considered to be objective and sufficient information for a particular context and/or problem requires the certification of social networks or user communities to legitimate its authority and transfer. Consequently, in order for scientific information to be influential, its value is negotiated and interpreted by a user community.  The focus on getting the science 'right' ultimately leads to a power politics where divergent actors and interests select and interpret their own form of 'right' science. Yet taking a social realist approach and focusing entirely on developing legitimate processes with the perceived competency of experts can become susceptible to collective bargaining and interest negotiation likely leading to power politics as well. A proposal for a pragmatic way of dealing with these problems, is to problematize the notion of realism altogether and to recognize scientific activities and scientific content as socio-technical hybrids. Scientific inquiry is made up of social, conceptual and material processes. Selections and interpretations are dependent on the observer, practitioner and analyst and are therefore imbued with underlying social and cultural commitments. In the policy realm recognizing that scientific information is a hybrid or socio-technical construction (neither entirely rational nor entirely social) transfers the focus from what information is 'right' to what is 'appropriate'. Determining what is appropriate in the public sphere therefore is not only a project of knowledge producers but includes knowledge users as well.  It follows that if the user community negotiates the 'fit' of scientific representations and interpretations at the science-policy interface, than this community also has a considerable hand in stabilizing scientific claims. The constructs of policy and science are not "structural necessity or faceless rationality" (Gieryn 1999, 89) as represented through the 'truth speaks to power' or 'power speaks to truth'. Instead in practice the boundaries at the science-policy interface undergo continual negotiation between and within both communities at the science-policy interface (Litfin 1994; Jasanoff 1995). This negotiation either expands or constricts the authority of science depending on the collective needs of the policy or 'user' community (Jasanoff 1995). The co-production of knowledge and power occurs in what are typically tacit negotiations at this interface. Exploring the contingent understanding of both policy and science as constructed by discursive and rhetorical strategies within a particular context forces a reconceptualization and reanalysis of the science-policy interface.  38  3.4 Boundary Work, Boundary Objects and Boundary Organizations  A growing literature is emerging on the ephemeral and idealized boundaries between science and policy (Latour 1987, 1993; Gieryn 1983, 1999; Guston 1991, 2000; Miller 2001). This literature concentrates on 1) the rhetorical boundaries between science and policy, 2) ways to stabilize them in the heterogeneous non-scientific community and 3) ways to ensure that the in situ negotiation at the science-policy interface contributes to relevant information that remains credible and legitimate relevant. Through what Gieryn calls 'boundary work' (1995, 1999), scientists or actors employ rhetoric to construct boundaries between different types of science and scientists, and between science and non-science. This rhetoric varies and is contingent on the context and the audience, which it is trying to convince of its social value.  33  Gieryn  (1983, 1999) argues that definitions of good science and bad science are ambiguous, flexible, historically changing, context-dependent and generally accompanied by considerable contestation. Yet boundaries are stabilized to stave off threats to the cognitive authority of science, such as pseudo-science or politics (Guston 2001) and to protect the status of science as a rational, objective and neutral epistemological tool (Gieryn 1999). He argues that,  Nature typically (not always) becomes beliefs-about-nature or accounts-of-nature when the focus is on facts as collective suppositions - suppositions that are consensually stabilized over time and space via through contingent processes of laboratory tinkering, or inscription making or ally enlisting (Gieryn 1999, 341). What this suggests is that particular facts relevant to an actual concrete problem or issue are collective suppositions agreed to by a community. These suppositions are simultaneously based on beliefs about nature and are stabilized through the contingent processes that contribute to the stabilization of claims such as ally enlisting. What is of interest then, is what community (or which communities) have cognitive authority to consensually stabilize suppositions (or premises) about nature? Constructivist approaches in the social studies of science break down essentialist representations of science to reveal the contingent, flexible and multivalent commitments involved in scientific practices and the production of scientific claims.  For instance, Gieryn (1999) reveals how in the late 19* century, Tyndall used different rhetorical devices to carve a role for science between religion and mechanics. The rhetorical strategies used to separate sciencefromreligion focused on empiricism, observable facts and practical utility rather than on mere religious or poetic understandings of the natural world. However the rhetoric shifted when differentiating science from the mechanical and technical trades. The virtues of science then focused on the theoretical and abstract, moving beyond material facts into invisible laws and reason. Science was seen also as a noble and cultured tradition distinct and quite separate from the brute guild of mechanics. Tyndall's efforts to separate the boundaries of the scientific from the religious and mechanical reveals the "interpretive flexibility" in the rhetorical strategies of science to remain distinct and separate from non-science. Gieryn (1999) argues that science is only characterized in relation to non-science. 3 3  39  Science studies reveal the ways scientific institutions and communities are not discrete social worlds acting independently from society, culture or politics. Instead, this seemingly discrete institution has implicit social and cultural commitments that act to produce and reproduce social power in mutually reinforcing ways (Jasanoff and Wynne 1998; Shackley and Wynne 1996). Science is not devoid of political values nor is politics (of rationality) devoid of some form of scientific legitimation (Gieryn 1983, 1999; Guston 2001). Instead they are a reflection of one another where through norms and practices, mutual arrangements between knowledge and power are created and recreated (Jasanoff 1990, 1996, 2003; Latour 1987, 1990). Jasanoff and Wynne (1998) argue that bureaucratic and scientific cultures cooperate in the simultaneous production of knowledge and social order leading to the systemic coproduction of information (also see Latour 1987 and Jasanoff 2004). Guston asserts that, "By clearly portraying science as it is practiced, constructivism can recover the human face beneath science's rationalist mask" (1999, 89).  3.4.1 Boundary Objects: Negotiating Representation The concept of boundary objects, introduced by Star and Griesemer (1989), provides a more nuanced scale and approach to examining the internal instability and the negotiated interface between science and policy communities. Boundary objects are shared representations that have both considerable purpose and meaning for two or more social worlds. Each world draws on repositories of information, ambiguous symbols and inscriptions that are used as "good enough" roadmaps, that rely on a level of abstraction taken from both concrete and theoretical data but delete local contingencies (Star and Griesemer 1989). Boundary objects are "both plastic enough to adapt to local needs and the constraints of the several parties employing them, yet robust enough to maintain a common identity across sites" (Star and Griesemer 1989, 393). In the negotiation of boundary objects there are no transcendent boundaries for cognitive authority - each side draws a different map to create science as a distinct ontological preserve and to claim legitimate and authoritative representation (Gieryn 1989, 341). In other words, boundary objects are standardized representations that are collectively negotiated and stabilized so that common representations can translate across the divergent social worlds and satisfy the informational requirements of each. A boundary object involves a process of stabilizing representations by maximizing both the autonomy and communication of divergent social worlds. What becomes critical in producing boundary objects are the underlying processes that encourage the interaction, exchange and negotiation across divergent social worlds in order to achieve this common representation while also preserving the role and professional identity of each community involved.  40  3.4.2 Boundary Organizations: Managing the Science-Policy Interface The identification of boundary work and the negotiated interface between science and policy communities have led researchers to examine ways of reducing possibilities for political contestation at the interface. Guston (1999) combines the sociological approach of 'boundary work' that demarcates science from non-science, with the political economic 'principal-agent' theory' to develop complementary theoretical insights into the ways that science can become stabilized in the policy sphere. He argues that the sociological detail of boundary work discourages scientists from overstylizing facts while the structure of principal-agent relations formalizes what may otherwise be a diffuse activity of science moving into the policy sphere. He argues for a new theoretical combination referred to as "boundary organizations".  A successful boundary organization will succeed in pleasing two sets of principals and remain stable to external forces astride the internal instability at the actual boundary. The success of the organization in performing these tasks can then be taken as the stability of the boundary, while in practice the boundary continues to be negotiated at the lowest level and the greatest nuance within the confines of the organization (Guston 2001,401). A boundary organization therefore maintains the appearance of stabilizing the boundaries of science in policy from the outside while in practice, internal to the organization these boundaries are continually being negotiated "at the lowest level and greatest nuance".  34  A boundary organization provides 1) the  opportunity for the creation and use of boundary objects, 2) involvement of participants from both sides of the boundary, including professionals who serve a mediating role, and 3) exist on the frontier of two relatively different social worlds of science and policy, but have distinct lines of accountability to each. Guston's (2001) focus on intermediary organizations that involve participants on both sides offers important insights about how to manage the science-policy interface in order to ensure that each social world gains in order to maintain accountable lines to each community. In other words, balance is maintained where each social world is able to construct the boundary between their enterprises that is favourable to their own perspective (Guston 2001).  In this framework, an intermediary organization lies between science and policy. Guston (2000) uses the National Institute of Health (NIH) as an example of a boundary organization that lies between the Ministry of Health and scientific researchers. The NIH pleases two sets of principles; it encourages scientists to patent their research ideas, while simultaneously ensuring the Ministry is provided with progress reports on scientific productivity. Another example is the Office of Technology Transfer (OTT) (Guston 1999). The inclusion of Democrat and Republican scientists in the selection and interpretation of scientific information ensured controversy and negotiation was maintained within the confines of the organization thereby reducing the potential for adversarial conflict and politicization in the public sphere. The OTT generated significant trust among policymakers for the impartiality and legitimacy of its scientific documents, but was eliminated by thefirstBush Administration. 3 4  41  Boundary organizations offer the appearance of consensus and scientific stability, while also providing space for considerable instability and negotiation internal to the organization (Guston 1999; Cash 2001). In this way, Guston's (2000) analysis identifies two science-policy interfaces; one external, which has received considerable attention among researchers and one internal, that produces boundary objects which has received much less attention (discussed further below). At the former scale of analysis, boundary organizations increase the legitimacy of science by minimizing the politicization among external forces. Though this theory focuses on ways to connect long-standing science and policy institutions in the US, it still suffers from the problem of linking the relevant information with the needs of the user or policy community. At this scale, boundary organizations do not overcome problems with a unidirectional model of science in policy (Miller 2001) and making scientific information policy relevant as well as politically legitimate.  3.5 Boundaries in the International Sphere  Applied to the international sphere, the negotiation of science-policy boundaries becomes far more complex with scientists from various socio-political and cultural contexts aggregating and interpreting scientific information based on different disciplinary goals and values, and different cultural frameworks. In this setting the diversity of international governments and stakeholders are even more heterogeneous. Once science moves into the policy sphere, interests, power relations, reputations and scientific uncertainties become resources used to persuade decisions one way or another (Wynne 1989). Trust among diverse actors is required and can only be established by stabilizing the boundary at the interface through social processes of inclusion rather than exclusion (as revealed through the concept of boundary organizations). Yet it is not enough to focus on getting the science 'right', especially in the international sphere where the multitude of different value frameworks could ensure that the concept of 'right' scientific information remain under critical and constant scrutiny and negotiation.  In the late 1990's the Global Environmental Assessment Project (1997) analyzed various multi-lateral environmental assessments to compare characteristics that contributed to each ones relative success (or lack of). The outcome of this project was that considerations such as: 1) the social context in which the assessment is operating and the ways issues are defined and framed, 2) that scientific information considered to be credible, legitimate and salient was more likely to influence policy responses, and 3) 35  explicit efforts to protect the boundaries of the science and policy communities at the interface or in other  For the purposes here, credibility refers to the scientific adequacy of technical evidence and arguments; legitimacy refers to the perception among participants that the information has considered divergent values and beliefs, has been unbiased in its conduct and fair in its treatment of opposing views and interests; and salience refers to the relevance of the assessment to the needs of the decision-makers. 3 5  words to perform 'boundary work' contributed to the overall success of environmental assessment processes (Clark et al. 1997). This research stresses the importance and the difficulties involved in producing policy relevant scientific information that remains credible and legitimate while also preserving the professional identities of the science and policy communities.  The understanding that information is negotiated and co-produced at the science-policy interface and is the result of a social process rather than unitary and universal scientific content provides some major insights into the ways that science in international regimes can be investigated. Borrowing from the concept of the socio-technical hybrid where the subject and object are interpenetrating forces (Latour 1987), Jasanoff (2003) adopts an idiom to describe the ongoing interaction between science and policy. She describes this as co-production where the natural order and the social order recreate one another in mutually reinforcing arrangements. The idiom of co-production has been used to account for the mutual ordering between science and policy. In the book States  ofKnowledge: The co-production of science and  social order, Jasanoff (2004) highlights co-production as a way of interpreting and accounting for complex phenomena so as to avoid the strategic deletions and omissions of most other approaches in the social sciences (Jasanoff 2004). Co-production is shorthand for "the proposition that the ways in which we know and represent the world (both nature and society) are inseparable from the ways in which we choose to live in it" (Jasanoff 2004). What is highlighted here is the need to design processes that manage the intersection between social worlds that include a broader peer community in negotiating relevant and appropriate scientific information. This type of science-policy interaction has been identified as an under-estimated element of design in international assessment processes (Farrell et al. 2001).  A defining aspect of mandated science is the recognition that the decisions and/or the documents that result are neither entirely scientific nor entirely political. It is the proposition here that because institutional arrangements are socially constructed (Skodvin 1999), as are categories of what constitutes science and non-science (Gieryn 1985, 1993), there may be more fruitful ways of constructing and designing science-policy processes that recognize the social processes and the context-dependence or situatedness of scientific practices. Processes are needed that not only manage science-policy interaction as in the boundary organization but also manage it in a way that recognizes the interpenetrating and inseparable forces of these two social worlds while remaining credible to the scientific community and legitimate to the policy community. Science-policy communities interact whether they are managed or not. Therefore a process that manages and facilitates science-policy interaction and interpenetration may be more successful at producing policy relevant scientific information. Without systems of global governance or entrenched science-policy relationships, the international sphere and in particular global  43  environmental assessment provides an ideal opportunity to investigate possibilities for increased sciencepolicy interaction at the global scale.  3.6 Developing the Imbued Meaning Framework  While both science and policy communities have their own discourses, values, goals, and knowledges which are necessary to maintain as credible and accountable to each set of practices and processes, it is critical in matters of environmental policy, to understand that both are interacting simultaneously in framing and describing problems. With the advent of global environmental change it is not enough to say that science is too uncertain, nor is it sufficient for nations to maintain old forms of diplomacy where interest realization and pursuit of uncoordinated goals and objectives ends in deadlock, disagreement and non-compliance. Both communities are required for the simultaneously normative and technical dimensions of global environmental change.  The Imbued Meaning framework was constructed as a pragmatic attempt to use the approach of the social studies of science that focuses on examining processes in the production of scientific information in order to tease out and address some of the problems introduced above about how scientific information becomes mandated, extrapolated and used in the policy sphere.  This framework focuses on ways to  improve the trust and reliability in scientific knowledge as it moves from the scientific sphere into particular policy/social contexts. The unidirectional and interactionist model of science described below are similar to what Robinson (1992) refers to as classical and critical views of science. The initial point of departure is an understanding that a unidirectional model of science is not a useful way to think about the science-policy relationship. Science-policy literature shows that increased spending on science and attempts to reduce scientific uncertainties is not necessarily fostering better policy decisions but may in fact be contributing to increased policy deadlock (Jasanoff 1997; Sarewitz 2000; Sarewitz and Pielke 2000). Although science contributes enormously to identifying and describing problems, numerous examples suggest conceptualizing its contents in this way may not lead to desired or desirable outcomes (see Winner 1986; Wynne 1982; Robinson 1992). Another intermediary model has emerged as a response to increasing public dissatisfaction and scepticism toward policy decisions and the use of science in decision-making that is assessed separate from its policy context (Robinson 1992). This new model recognizes that the policy context is important for generating policy relevant science. It appears in participatory models of science where there are attempts made to focus on more effectively communicating science to the policy sphere by expressing information in assessments of risk and inclusion of scientific uncertainties (Robinson 1992). Although these participatory processes aim to become policy relevant and more effectively communicate to the 'users' of science (or stakeholders in a  44  problem), these processes "retain the role and status of science as providing value-free knowledge" (Robinson 1992, 249) and therefore offer rhetorical rather than real inclusion of the user community. The Imbued Meaning framework is an attempt to articulate an interactionist model of science-policy, or what Robinson (1992) refers to as a critical view, in order to derive credible and legitimate processes for constructing policy relevant scientific information. The Imbued Meaning framework is a normative prescription that argues that mandated science should be "more integrated into the policy context, more contextual and openly value-laden, less oriented to mastery over natural and social processes, and more accessible to the public [and policy community] at large" (Robinson 1992, 249). The main insight is that the science-policy interface includes and should include both knowledge producers and knowledge users in a structured process of interaction and negotiation in order to generate policy relevant scientific information.  The Imbued Meaning framework establishes an interactionist model that focuses less on getting the science 'right' in global environmental assessment forums, and instead focuses on producing reliable and meaningful policy relevant scientific information that remains credible and legitimate to both scientific and political communities. Each element of the framework individually offers a path to produce scientific information that is credible, legitimate and policy relevant. In particular the ability to integrate the user in problem framing and to reappraise information at some future time contributes to greater science-policy interaction and is the basis for producing more meaningful policy relevant scientific information. Five process-related elements are proposed in the Imbued Meaning framework. These elements are highlighted in Box 3.2. The elements of the Imbued Meaning framework considered together reconceptualize the status and role of science in the policy sphere as a series of processes that facilitate interaction. This is done to ensure that the policy sphere pushes back and exerts a transparent and managed pressure on the scientific realm to ensure that knowledge users can have influence on how scientific information is framed and represented while also contributing to normative judgments about which information is considered adequate and appropriate for the policy context.  45  Box 3.2 The Five Elements of the Imbued Meaning Framework Weak Co-production (Points 1-3) 1. A managed interface that creates space for formal science-policy interaction 2. A bi-directional exchange of discourses and materials occurs across the science-policy interface. 3. A formal negotiation of the final message and representation of the scientific information.  Strong Co-production (Points 1-5) 4. Includes the 'user' community in problem framing. 5. Reappraisal of information at some time in the future  First, the concept of a boundary organization is used to highlight the need to stabilize the identities of science and policy communities among external forces. In order to retain the credibility of the scientific assessment process each community must be perceived as pursuing its own objectives, norms and the practices. The identities of each community must be perceived as discrete. This is particularly important for ensuring the credibility of the scientific process. Any infiltration by the policy sphere can be perceived as the introduction of partiality, thereby reducing the legitimacy of the documents or recommendations in the policy sphere. However discrete territories and jurisdictions do not contribute to legitimate, relevant or appropriate scientific information. Instead a formal interface is required where the two communities converge. Managed and agreed to processes and protocols are required in order to facilitate discursive interaction between these two communities. A n interface is necessary to ensure that key and critical findings are translated into the policy sphere in ways considered acceptable and appropriate to the international policy community.  Second, the unidirectional model of science for policy is illusory. It leads to either the policitization of science or the tacit transfer of values and assumptions in the scientization of policy. With increased recognition that science has associated underlying assumptions, social and cultural commitments and uncertainties, it becomes increasingly clear that once these tacit aspects move beyond the goals and interests of the scientific profession into the non-scientific sphere scientific information needs to be interrogated among an 'extended peer community' (Functowicz and Ravetz 1992). Interrogation of the sufficiency, adequacy and appropriateness of information is not only the work of scientists and experts, it also occurs at the interface with the user community. In order to derive policy relevant scientific information, this interface must ensure that the mutual exchange and negotiation of discourses, values, knowledges and interests between communities is encouraged. Therefore materials, discourses and  46  values need to be transferred across the boundaries of the community in a bi-directional rather than a unidirectional way.  Third, although the focus here is on the process, the content is equally relevant in mandated science. It may be valuable to exchange materials and discourses and facilitate and manage interaction between the two communities. However the problems identified in the mandated science framework is that conflicting value-frameworks lead to contestation and disagreement over appropriate selections and interpretations of information. The final representation of the science cannot merely be transferred or 'translated'. It is negotiated at the interface whether managed or not. The third element of this framework is a managed interface where science-policy interaction is encouraged and where formal procedures for conflict resolution are made available. The inclusion of the user community in formally negotiating what is policy relevant scientific information encourages discursive interaction between these communities in a credible way. Policymakers do not have the accreditation to determine what is and is not credible scientific information nor do scientists have the authority to determine what is policy relevant. In order to produce policy relevant scientific information that is credible and legitimate, a formal negotiation is required where inputs from both communities. Therefore the negotiation toward a common representation that is agreeable and agreed to offers a formal and credible way for science and policy communities to cooperate and to co-produce policy relevant scientific information.  Fourth, for strong co-production to occur, the user community needs to be included in the framing of the public policy problem and in establishing the information required to address it. Scientific research can be misguided with regard to real-world problems and the needs for addressing them. Rather than having scientists assume what those needs are, in areas of mandated science it may be more advantageous to develop a process where a range of 'users' with interests in the outcome of policy decisions are included in framing the problem (Irwin and Wynne 1996). For instance, in the case of the Windscale-Sellafield Inquiry it was found that if sheep farmers had been included in the framing of scientific questions about location and impacts of radioactive fallout, the scientific process would not only have been more efficient but more legitimate and relevant (Wynne 1989b). This inclusion of users in the framing provides an opportunity for knowledge producers and users to interface and to determine the informational requirements to address.  Fifth, these institutional arrangements should not be viewed as static instead they should be constructed as dynamic and ongoing. Policy relevant scientific information should be reviewed and revised in light of new scientific information and new policy concerns. In this way the provisionality of science is highlighted, and the paths forward are neither technocratic nor static but may change over time.  47  Moreover, Eckley (2002) finds in her research on policy negotiation that policymakers are more predisposed to agreement when they can agree that the information will come up for reappraisal at some point in the future. Including the user in the framing, interrogation and negotiation of information may be more palatable if the information can be changed at some predetermined point in the future. For strongly co-produced policy relevant scientific information, all four elements of the Imbued Meaning framework can be applied and reappraised iteratively (the fifth element) to deal with the problem at hand.  36  3.7 Implications of the Analysis  The permeability of the boundary between science and policy has shown that rather than having separate and essentialized domains, science and policy have boundaries that are negotiated in situ on individual topics and domains (Jasanoff 1990). In other words, negotiations are generally taking place at the interface whether they are managed or not. Yet it is important to note, unless one maintains a positivistic and unidirectional understanding of science, the identification of social processes does not undermine the usefulness of scientific products. Shaw and Robinson (2004) examine the ways that positivist accounts of science in the public sphere have the paradoxical effect of derailing scientific authority in the policy sphere. In potentially urgent and irreversible policy problems, such as those posed in environmental policy problems, the demand by policymakers for robust and consensually agreed to technical information, may take decades to produce. In the interim scientific information with considerable uncertainties can be used as a political resource to legitimize policy inaction. Should the reduction of uncertainties be the main requirement of science and scientists in the arena of science-policy?  In this context, the epistemological shift to a constructivist understanding of science offers a new way of thinking about and approaching problems of credibility and legitimacy of science in the policy sphere. The focus on the practices and processes of science shift the analytical perspective from that of idealized theory to that of an understanding of idiosyncratic, particular and local practices.  The Imbued Meaning  framework asserts that the management of science-policy interaction, using processes considered credible and legitimate to both social worlds, can not only contribute to a negotiation of what is considered to be adequate, relevant and appropriate scientific information for a particular policy problem but can, through discursive interactions, imbue that information with normative meaning.  This framework is intended to derive policy relevant scientific information that recognizes the hybridity of all information while also recognizing the need to preserve the identities of the Science and Policy institutions and the perceived impartiality, credibility and legitimacy of the information produced. However, it also works to challenge false dichotomies at the interface where, in this view, risk assessment/risk management, technical/normative and fact/value distinctions are seen to be interpenetrating and inseparable (Latour 1993). 3 6  48  Science is recognized, as an important social project comprised of standardized and rigorous methods. In the arena of mandated science its usefulness (rather than its authority) is about offering well-analyzed materials that have been collectively reviewed by specialists that are then negotiated for their value in the policy sphere. The Imbued Meaning framework encourages both of these critical processes of validation to be managed through an explicitly constructed assessment process.  ,  3.8 The Following Chapter The Intergovernmental Panel on Climate Change (IPCC) provides an ideal case study of mandated science at the international scale. The three iterations of the IPCC Assessment Reports (1990, 19996, 2001) have encouraged iterative appraisal and reappraisal of peer-reviewed climate information. With these iterations there have been concurrent attempts to increase the transfer of scientific information to the international policy sphere in a way that is both legitimate and policy relevant. The way the IPCC has adapted its processes and protocols in order to increase its credibility, legitimacy and relevance is of interest. In particular, in the Second Assessment Report (SAR) (1996), two procedural innovations were generated that created interfaces between the science and international policy communities; these were the Summary for Policymakers (SPM) and the Synthesis Report (SYR). The SPM was formally described as a way to extract information that may be politically important from the supporting material of the underlying Working Group reports (Bolin 1994). In the SAR, a W M O executive council decision to make more explicit links between the findings of the underlying W G reports and the objectives of the U N F C C C , with specific focus on Article 2 resulted in a Synthesis Report (IPCC 200Id).  In order to explore these procedural innovations in action, a constructivist methodology is used in order to develop insight into possible differences between this formal rhetoric and the actual practices. The next chapter describes the methods that were used to examine these differences.  49  C H A P T E R I V : Methodology  The methodological challenge for this type of research is to find a way to observe and analyze formal and informal science-policy interactions at the interface. The Intergovernmental Panel on Climate Change (IPCC) was chosen as a case study due to its formalized processes and practices that include the policy community in the development of the policy relevant scientific information. Access to the process was made possible by my supervisor, Dr. John Robinson, who was a Coordinating Lead Author (CLA) in the IPCC's Third Assessment Report (TAR) Working Group (WG) III during the time of my research. This information is constructed through a managed science-policy interface in the Summary for Policymakers (SPM) and the Synthesis Report (SYR) and its SPM. These specific and formal interfaces offer a location from which to observe and analyze interactions at the science-policy interface and to apply my theoretical framework, Imbued Meaning discussed in the last chapter.  The formal understanding in the IPCC was that scientific information needed to be more effectively communicated and transferred to the policy sphere leading to the development of the SPM and SYR (Bolin 1994). This interface was initially created to ensure that scientists transferred key policy relevant findings to the policy community and that governmental representatives were included to appropriately communicate relevant scientific findings (Bolin 1994). This representation suggests a unidirectional transfer of scientific information to the policy sphere. Therefore it also provides a compelling place from which to assess the Imbued Meaning framework. The SYR follows a similar process however at a greater scale of aggregation and synthesis. The SYR is another level of assessment that encompasses all three WG's from the T A R . It is framed by policy relevant questions, approved by participating governments to the Conference of the Parties (COP) and follows essentially the same process as the SPM. The methodological issue then becomes how to observe and analyze the informal practices and negotiations in the SPM and the SYR.  4.1 Particular setting The case study approach delineates the limits of a study by temporally and spatially bounding the area of research. The SPM and SYR processes are embedded within the larger mandated science assessment process of the IPCC.  This study identifies participants in the SPM and the SYR as part of a particular  scientific culture with unique practices. A n ethnographic approach was used to observe, describe and explain the particular cultural activities involved in these forums. This enables observations and explanations to be made about specific details and contributes to a nuanced understanding of the what things go on in that community; why and how these things occur; who among the community members  50  take part in these activities and behaviours and social forces that may bind together members of this community. O f particular interest are the ways information is constructed, crafted and reviewed in this scientific culture and how this transfers across boundaries into the international policy sphere.  4.1.1 Why the IPCC was chosen The creation of the IPCC in 1988 constituted a watershed in the scale and scope of international science assessment. Since then, the IPCC has issued three Assessment Reports (1991, 1996, 2001), each consisting of three volumes, amounting to thousands of pages, and involving the participation of thousands of experts around the world as authors and reviewers in the assessment process.  37  In an effort  to draft "policy relevant but not policy prescriptive scientific information" (IPCC 1999), theTPCC has developed two innovative methods of synthesizing scientific information: the SPM to summarize the results of its detailed working group volumes (used in all three assessments), and the use of policy relevant scientific questions (PRSQ) to structure the final Synthesis Report in the 2001 assessment. The existence and development of the IPCC in its three iterations since 1988 provides a powerful case study for the way science has been used in support of the policy process.  M y research focuses on the negotiation of the WGUI SPM of the T A R , the drafting of SYR and the negotiation of the SYR SPM. The data was collected between February 2001 and September 2001 in which time two SPM processes were approved and published shortly thereafter. The SPM process provides a particular and manageable environment where there exists a fuzzy boundary between the social worlds of science and policy. It is therefore identified as fertile ground to attempt to understand the ways that these worlds intersect and are managed through formal processes and protocols in ways that remain credible and legitimate to participating scientists and delegations. There have been studies done on the history of the IPCC (many of which are used in my own analysis) (see Agrawala 1999a, b and Boehmer-Christiansen 1994a,b in particular). There have been studies on the content of climate science either used or produced by the IPCC (see Shackley and Wynne 1995, 1997; Shackley et al. 1998; van der Sluijs et al. 1998). To my knowledge there has been little to no research done on what constitutes policy relevant scientific information and the implications of this overarching type of mandate for science in the policy sphere.  The three volumes of the 2001 Third Assessment Report, and the Synthesis Report were published by Cambridge University Press under the title Climate Change 2001. The full text of each report and the various special reports can also be found on the IPCC website at www.ipcc.ch. 3 7  51  Although formal guidelines and protocols have been written to establish the working relationship between the science and policy communities (IPCC 1998, 1999), rules do not necessarily govern behaviour and action. This is the difference between the naive realism in the philosophy of science, which focuses on the rules of conduct that produce scientific knowledge to provide for more legitimate information, and constructivism in the social studies of science where it is argued that methodological criteria may have limited bearing on actual practices (Lincoln and Guba 1985; Guba 1990). Whether or not these rules of action are followed and, if not, where the procedural transgressions lie is of particular interest. It is through heterogeneous interpretations and social processes that useful information can be examined about the diffuse social worlds of science and policy. In the same way that Latour called for a study of 'science in action', I feel it is appropriate to produce a study of the IPCC in action in order to better understand what the actual practices are in this international forum. How information is represented and transferred is the particular focus of a constructivist methodology. Gieryn (1999) states that rather than accounting for boundary work between science and policy communities it may be more valuable to understand the constructed nature of these boundaries and rather than essentializing the boundaries it may be more fruitful to look to how and why they are constructed in particular cases.  4.2 Case study The research can be summarized in one sentence: A  case study that describes and reflects on the process  and outcomes of the SPM and SYR innovations of the IPCC to understand the activities and practices of the science-policy interface, the close working relationship between these two communities, and the processes that maintain credibility and legitimacy of the product generated by both social worlds. From a methodological perspective, this thesis statement broadly captures all the elements of the case study method used, which include participant observations of the activities and practices involved in these forums, the analysis of the formal management of this working relationship through document analysis, and the ethnographic methods of interviews to develop an interpretive analysis of which processes and protocols provides credibility and legitimacy to these processes. A l l methods are situated in a particular research context and therefore generalizations cannot extend to other contexts but are instead used to explore theoretical possibilities. The following chapter describes how these methods manifested themselves in the research, argues why they were chosen and describes some the theoretical background for using them.  There is a broad range of case study methods from which to choose. The methods chosen for this research are diverse and in some sense obvious. For instance an experimental strategy was inappropriate for research into the activities and practices of the SPM and SYR since I have no control over the behaviour of the individuals involved in the research setting nor do I have a control group for comparison 52  (Miles and Huberman 1994). The issue is contemporary and therefore historiographical or archival methods were not relevant. Surveys may have been useful to develop a discursive understanding of how scientists perceive the SPM and SYR processes and their involvement in these science-policy innovations, however participants perceptions are not entirely reliable to garner an understanding of process. Three diverse approaches were taken to develop interpretive strategies for understanding the IPCC process and in particular the SPM and SYR processes without having to rely entirely on any one interpretation (whether the formal rhetoric or the perspective of a rogue scientist). These were document analysis, participant observation and participant interviews.  The strategy of individual case studies is not to collect samples or generalizable facts but is instead to look into a particular and unique context in order to describe characteristics, activities, participants and practices. Case study methods attempt to answer the 'how' and 'why' questions "involved in particular research settings where questions are being asked of a contemporary set of events over which the investigator has no control (Yin 1984). The research context under investigation is the IPCC in action. Based on the nature of this research, the case study strategy seemed most appropriate. The ethnographic techniques of participant observation and participant interviews are used to provide a rich, more nuanced understanding of the research context, its social organization and the role of individuals within it (Yin 1984; Miles and Huberman 1994). Document analysis provides supplementary data that can be analyzed and cross-examined in conjunction with what is observed in the process and what is said about the process.  4.3 Using a Constructivist Approach It is through formal and informal processes that insights can reached about what judgments, assumptions and commitments are made and where in scientific inquiry. For instance, Shackley et al. (1998) directly observe climate scientists and modellers as they construct general circulation models (GCM). These social studies of science (SSS) researchers perform a constructivist investigation into the social processes involved in constructing such models and find that heterogeneity and contestation exists among purist and pragmatic scientists. They find that scientists have differing interpretations about the use of flux adjustment in the coupling of atmosphere/oceanic G C M ' s that are based on their assumptions about what are considered to be sufficient and adequate methods, techniques and data for the policy sphere (Shackley et al. 1998). Constructivist accounts can provide nuanced accounts of the social processes and commitments that influence the representation of scientific information.  Phenomenological and constructivist researchers argue that the researcher is unable to find absolutely true or objectively correct interpretations due to the inability to conclusively verify and falsify particular  53  claims (Shapin et al. 1985; Latour 1987). Constructivist studies suggest that the natural world is never discovered and verified though scientists and users of science often represent it in that way after the fact. Instead scientific information is context-dependent, ambiguous and contingent, and the sufficiency and adequacy of claims are negotiated rather than verified. Constructivist approaches highlight the valueladeness of all explanation. In keeping with this approach it is important for me to make transparent the explicit constructivist assumptions that underline my research. Using a constructivist approach both as a method of description and theoretical explanation forces me to situate myself as the researcher, and my own research within this tradition. Considerable reflexivity is required to situate the researcher's theories, assumptions and approaches in relation to the research in question. This has become a standardized treatment of the contingent nature of social research and efforts will be made to convey appropriate methods, procedures and limitations below.  4.4 Research Methods Participant observations were taken from three meetings, supplemented with a series of semi-structured interviews and document analysis. All data were collected over the course of three different meetings. The first was the 17 Session of the Session of the Panel of the IPCC held in Accra, Ghana in March th  2001. The Session of the Panel "refers to a series of meetings at the plenary level of the governmental representatives to the IPCC" (IPCC 1999). This meeting focused on the acceptance of the underlying report from WGIII in the T A R and the approval of the WGIII SPM. The second was at the scientific author meeting held in Annapolis, Washington in June 2001. This was a meeting of the C L A ' s involved in drafting the SYR report and its SPM. The third was the 21 Session of the Panel held in Wembley, U K st  where the SYR and its SPM were accepted and approved by delegations. Recorded observations and a collection of internal IPCC documents were taken from three different meetings attended.  In total I attended 19 days of the IPCC process over the course of one year. The IPCC Chairman, Dr. Robert Watson, formally approved my participation largely on the basis of my alignment with WGIII C L A Dr. John Robinson (see Box 4.1). To my knowledge I was the only researcher authorized to observe the Panel Sessions (there are many formal observers such as NGO's and industry however). I am formally listed as a researcher in the Participant List of the SYR and I was the only authorized observer in the L A meeting held in June 2001 in Annapolis, W A .  4.5 Sources of Data These three meetings were observed with different scales of observation, description and analysis in mind. Broad understandings of the science-policy interface and boundaries between these two social 54  worlds were examined; areas of consensus were compared to areas of heterogeneous interpretation and controversy; informal interactions between scientists and experts were observed; as were national negotiation styles; and the participation, reputations and authority of individuals. These were the primary areas under investigation in all areas of data collection, reduction and analysis. It is believed that exploring these various aspects of the process led to a more comprehensive understanding of the ways credibility and legitimacy are formulated in the SPM and SYR processes.  Box 4.1 A Research Anecdote  I surfed into this participant observer position on the coat tails of IPCC WGIII Coordinating Lead Author and my supervisor, Dr. John Robinson. His trustworthy reputation and his interpersonal network within the meetings were implicitly transferred to me. This helped to establish a level of trust within and among many of the scientists. It also helped me to have the confidence to approach many of the key players who otherwise may have claimed they were too busy in the meetings. In fact, my debut at the T A R WGIII Session of the Panel held in Accra, Ghana, I relied heavily on Dr. Robinson for introductions due to the fact that I lost my voice. This is a researchers' nightmare. All the preparation, funding and resources used to get to this place in order to observe and perform interviews and I lost my voice from the air conditioning my first night in the hotel. This was definitely not part of the plan and was, in fact, very embarrassing yet it may have proven to be a benefit over the course of the 9-day meeting. It allowed Dr. Robinson to introduce my research, and me while at the same time letting the scientists know that I would be interested in interviewing them once my voice was back. This provided me with sympathy points that became quite useful. A number of interviews were performed after being approached by certain key players who were concerned about how I was feeling and whether my voice was back. (I began interviews as soon as the air coming from my voice-box was audible. A number of the interviews were done with my horrific voice captured on tape). I seized these opportunities as soon as they emerged. This included the chairman, Dr. Robert Watson. As soon as he asked how my voice was, I was able to seize the moment. M y reply was "My voice is finally audible, although slightly grotesque. I have a number of questions I'd like to ask you. Could you withstand this for a half hour of your time?" Oddly enough, this on-the-surface horror for a social researcher was a blessing. I met more people, developed relationships faster and was more recognizable than I would have been otherwise.  55  4.5.1 Participant Observation Process observation focuses on the group dynamics and processes which people use to accomplish tasks. Observing the roles assumed by different members within a group, the interactions between participants in the process and the means of decision-making utilized reveals the kinds of institutional structures and formal protocols this particular culture is working under (McKinney 2000, 49). M y desire was to understand the particularities of the research context in order to recognize the areas where practices and behaviours deviated from the formal norms created in the SPM and SYR processes. In my note taking, I focused on the transgressions or infractions in the processes that constructed the SPM and SYR (what information was being discussed, what arguments were being made) and the catalysts (who is participating, what are the demands being made). These observations were then used in attempts to interpret the meaning of these transgressions (what motives, understandings and fears contributed to the fracture in process).  In cultural settings, patterns become internalized and can go unchallenged or unrecognized without  in situ  comprehension of the associated meanings. Through participant observation the researcher is able to discern the social norms and patterns within the particular research context (Yin 1984; Lincoln and Guba 1985; McKinney 2000; Thomas 2003). Through participant observation she can learn overt, explicit cultural behaviour as well as tacit rules and associated behaviours (McKinney 2000). However this technique also requires considerable reflexivity on the part of the observer, 1) to identify and expose the taken-for-granted activities through systematic observation and 2) to ensure that the explicated meaning from these activities is made reliable through corroboration with interviews or other sources. The ethnographic tradition recognizes how participant observation and interviews supplement one another and expand the knowledge of the researcher through her interaction with the research community (McKinney 2000, 93).  The process of the SPM culminates in the Session of the Panel. This is organized similar to a U N conference where all participating U N delegations use consensus negotiation in order to approve the SPM and SYR information on a line-by-line basis. The approximately 150 delegates are organized in a series 38  of concentric semi-circles with the IPCC Chairman and Secretary at the front alongside the relevant lead author (LA) scientists and the particular technical support unit (TSTJ). The Second Order Draft of the SPM is introduced into this setting after it has gone through expert (Zero Order Draft) and governmental/expert review (First Order Draft). This draft is projected onto a 15-foot screen for all 192 nations to see as they begin to negotiate this draft line by line for final approval, using Microsoft.Word Governmental delegations are composed with different numbers of officials rangingfromone official to eight officials.  3 8  56  track changes. The C L A ' s (scientists that contribute to the writing, editing and presentation of the reports, SPM's and SYR) are responsible for the particular section of the SPM draft being negotiated and sit at the front of the room in order to answer questions about the text and to prevent incorrect or misrepresentative changes from being made by the delegations. This process reveals a close working relationship between scientists and policymakers that relies heavily on procedural rules to demarcate science from non-science. However this setting is populated by delegations strategizing and politicking in the corridors often with IPCC scientists. Scientists are encouraged to informally interact with the delegations in order to ensure that a broad and comprehensive understanding of the information is negotiated. The significance of the information under negotiation in the SPM is considered to be critical by both scientists and delegations whose time is spent drafting and crafting very particular text and messages through formal and informal 'corridor-style' interactions.  Hot spots: Conflicts in the contact groups  In a process where each nation is equipped with veto power over the consensus-based process, conflict is expected. The conflict resolution mechanism within these meetings was to send controversial text to smaller contact groups of the relevant government delegates and C L A ' s that ran during the breaks of the Panel Session and, at times when resolution was difficult, parallel to it. These contact groups were interesting and crucial to observe. This is where the science and policy boundaries were pushed to their limits with generally one to two nations utilizing their veto power to ensure that a scientific message was rewritten, re-emphasized or removed. In these informal sessions, a scientist chaired the meeting at the front of the room, where text was projected onto a screen and Microsoft Word track changes was used to rewrite text according to the deliberations and negotiations among delegates and scientists. In these contact groups, the formal layout of the Panel Session was reduced to an informal setting with individuals (both scientists and delegates) deliberating, positioning and arguing over appropriate wording and messaging. All participants in these sessions agreed to the principle that no changes are possible unless they can be justified in the underlying text, and at the end of the session, there had to be general agreement that this principle had not been jettisoned in order for the controversy to defuse and for approval of the text to occur in the larger Plenary. Boundaries were blurred in these informal venues where the negotiation was partly about massaging the scientific message in a way that remained credible and partly about distinguishing between policy relevant and policy prescriptive scientific information. M y written notes attempted to capture and understand the strategies employed to reach agreement in the contact groups (footnotes, words, change to passive voice, etc). However note-taking also has its limitations (speed, substantive issues, interruptions, etc.).  57  Direct observation has the advantages of a) providing information from spontaneous, unplanned, unexpected events, b) not requiring any special equipment and c) being amenable to difficult contexts (noisy or crowded settings) (Thomas 2003). However the observer's note-taking can easily be disrupted either by the writing process itself or through other random events (a participant engaging in discussion for instance). In my case, the speed with which the discussions and negotiations took place at times made it difficult to fully understand due to the complex and substantive nature of the information being discussed. This made it difficult to comprehensively record. I had considerably less history and finesse with the substantive issues and materials being discussed than every other person in the room. I used my own point form notes to sketch who was speaking and what informal processes and concessions are made. At points where clarity was required, informants were used to help with clarification.  Following the action: In-situ research judgments  I found that, as a researcher interested in observing the activities of the contact groups, I had a,similar problem as many of the smaller nations: how to choose which contact group to attend. Similar to onedelegate nations (many delegations had 5-8 delegates in attendance), with the capability of only one individual, it became impossible to attend all contact groups and meetings. Therefore decisions about where to locate myself became part of the selection process that delineated the particular areas of focus. As Y i n (1984) notes, the case study approach is based less on what techniques are chosen and more on what the researcher chooses to study. In my observations, I paid particular attention to the normative areas being negotiated, observed transgressions in the process, and the ways these were resolved. These choices are in  situ judgments that fundamentally determine the kind of data that were collected.  4.5.2 Interviews Rapport and representivity M y attendance at multiple meetings helped to develop trusting relationships. Scientists and delegates were used to seeing me included in both the formal sessions and in informal events and activities (dinners, drinks, etc.). Many WGIII scientists, Review Editors and Bureau members were used to seeing me at the meetings. Among them, many knew my research interests and my reasons for being there and would approach me with strong interest in my findings or with observations and contributions of their own. These informal interactions enhanced my inclusion in and understanding of the process.  58  Selection of interview participants  The objective for the interviews was to sample a selection of scientific perspectives from different disciplinary and national backgrounds. As an observer in the T A R WGIII Session of the Panel, I was able to access a number of key WGIII scientists and a number of Bureau members. In the Annapolis L A meeting where the PRSQ's were being addressed by 35 C L A ' s from all three WG's, many of the WGI and WGII authors did not know of my involvement. Yet because of the intimate size of the group a few approached to ask me what my involvement was. This was an opportunity to secure interviews, however it met with less success than in the WGIII Panel. Again, the SYR meeting in Wembley was not nearly as productive. Although I got approvals for many interviews in this process, the intensity of the meetings meant that the majority were not performed due to time constraints. It is worth noting that the heightened interview success in the WGIII Panel Session lead to a discrepancy in the representation of scientists interviewed from the three WG's. Although representativeness was initially desired, it is not required for ethnographic research. "The prime concern is with the  conditions under which the construct or theory  operates, not with the generalization of the findings to other settings" (Miles and Huberman 1994, 29).  This is not to suggest that my interview strategy was to wait for potential interview candidates to approach me. Anthropological theory suggests that this kind of self-selection may prove to be more of a hindrance than help. Those informants wanting to forge relationships and make statements to the researcher are often on the margins of the social organization under investigation (McKinney 2000). Or in the case of the IPCC, the potential informants could have been either highly critical or full of praise for the process and want to vocalize them. Recognizing this, I drafted a list of 15 scientists, Bureau members and delegates that I was interested in interviewing. I approached 12 of those 15. O f those 12, only 9 were interviewed due to scheduling problems. Although all scientists asked agreed to be interviewed difficulty came in scheduling convenient times. For instance a U K scientist, China scientist, US scientist and Saudi Arabia delegate all agreed to be interviewed however when it came down to finding time in schedules or approaching them in the corridors they were always involved in high-level meetings. The other 11 authors, 3 officials and 1 observer interviewed either approached me (as discussed above) or were participants I was able to access in the corridors.  This latter ad hoc interview approach was initially a concern but may have been a blessing in disguise. The fifteen scientists I was initially interested in interviewing were all high profile veterans of the IPCC process, meaning they had been involved in more than one of the three assessment processes. In retrospect, this may have biased my analysis as they all had history in the process and looked upon it favourably. In other words, these high-profile veterans of the process had fully conceptualized the 59  process and had prescribed justifications for why it was a legitimate endeavour that deserved their continual involvement. This is not to say that their justifications are necessarily the same but certainly there is a feeling that there is an IPCC "party-line" in the ways that the process is conveyed and described among those with a considerable history of involvement. Alternately, the corridor interviews randomized the selection process a little more.  How interviews were performed The interviews were semi-structured in design. As discussed above, surveys and questionnaires could have been used in this research design, however the benefit of one-on-one interviews is that they provide the researcher with greater flexibility in exploring the research context. Interviews, and in particular semi-structured or loose interviews, allow greater freedom for the participant to develop thoughts and perspectives. Rather than a tight question or structured interview design intended to discover participants' preferences among a limited number of options, the loose-question approach is used to elicit participant interpretations of a general nature (Lincoln and Guba 1985; Miles and Huberman 1994). This has two benefits: 1) it enriches the story for the researcher to access and 2) through probing, it limits problems associated with premeditated responses.  Twenty-four semi-structured interviews were conducted over the course of three IPCC meetings (See Table 4.1 below). These interviews fulfilled the U B C ethics review criteria. Each interview was recorded with a hand-held recorder and consent was given. I simultaneously took notes in order to supplement my understanding of what was being said for the analytical process and to provide back-up in case of technical failure. A question template was used that was compartmentalized into the four key areas of interest for my dissertation in 2001. At this point the four areas of interest framed the structure of my thesis: Chapter 1) Development of the IPCC (Overarching Questions), Chapter 2) Constructing the IPCC, Chapter 3) Structure and Process, and Chapter 4) Direction for Future Global Environmental Initiatives. Under each heading there were 2-8 questions (see Appendix A for Interview Schedule).  Questions were asked in an informal conversational style once again to elicit as much breadth and depth as the participant was interested in divulging. This meant that not all of the 21 questions were asked, nor were they asked in any particular order. Instead, the interview template was used as a reservoir where questions were elicited to fit the ensuing conversation. Although this meant that I did not ask exactly the same questions to each interviewee or participant, it meant that a greater breadth and depth of material was collected. The disadvantage of semi-structured interviews is that the comparability between interviews is reduced. Coding becomes slightly more complex as innuendo and nuance underpin  60  conversational style. However the semi-structured style of interview was more casual and I suspect, built a conversation, rapport and trust with the participant more than a linear and formal style of questioning. M y perception was that this interview style enhanced the quality of the material elicited. Narrative storylines are followed rather than a direct comparison of responses to the same questions.  61  Table 4.1: List of Authors, Governmental Officials and Observers Formally Interviewed National Formal  Interviewees  Representation  Position  WG Status  Scientist Head Clmi.ile Research and Seismology Department. Renal Netherlands Meteorological Institute ( K N M I ) 1 1 oils Baede  Netherlands  D c B i l l , The Netherlands  L e a d A u t h o r . l ' A R VVG1  Si moi A V s i w i li A s S O L lute  2 Terry Baikei  UK  Department o f A p p l i e d Flco.ionucs  Coordinating 1 cad  University o f ("ambudge  Author. 1A R W G I I I and  Cambiidge, U K  Environmental Consultant International Petroleum Industry 1 inironnientdl  Conservation  A s s o c i a t i o n (IIMI L \) i  I cnny  Bernstein  US_  Meluclicn \ J , L S / \  Reviewer. I A R WGIII  I'ro/CSiOt, l . m ironmental P o l i c y and Management G r o u p . Imperial C o l l e g e a n d D e p a r t m e n t ni A p p l i e d I conomics, 5  Michael Gruhb  Cambridge LniveiMt}  Coordinating Lead  Cambridge. U K  A .ili>-i  I \ R WullI  Senior Kcsean Ii Spa talis! Sustuunihlc Dt \clopmcm Cooi dinator, Risoe Centei o n rnergv. Climate  I ead Author.'1 A R  and Sustainable  WGIII  Development  Coordinating I ead f> K n M e n  IlaKnaes  Denmark  Roskilde. Denmark  Author, S Y R  D i r e c t o r of R e s e a r c h ( R N S . Centre International de Recherche s u r I T r n i r o n n e m e n t et l c Developpoment "  .lean-( harle.s H o u i c a d c  fiance  (CIR1 D )  Pans, Fiance  Coordinating I ead Author. I A R WGIII  Dun tor, Netherlands I-nvuonmental A s s e s s m e n t A g e n c v (R.1V M ) S B e i t Met<r  N e t h e i hinds  B i l t h o \ e n . The N e t h e r l a n d s  C o - C h a n . I \R  WullI  PlOh'SsOI. D i r e c t o i of I n t e r n a t i o n a l 1 nviioninent and Resource Policv Program. The Iletcher School o f L a w and Diplomacy 9. B i l l M o o m a w  US  I jfls I'mveisiiv  I j a d Authoi.F A R  M e d ford M a s s . ! U S \  WGIII  National Formal  Interviewees  Representation  WG Status  Position  Scientist President. Biomass Users Network C L N B I O B r a z i l i a n Reference  10 J o s e R o b o r t o \ l o i c i i d  Bia/.il  Center o n Biomass  Coordinating Lead  Cidade Lrmersitaria  Author. IAft W ( . I l l and  Sao Paulo. B r a z i l  Professor, 1 )ireetor o f S o c i o - L n v i r o n m e n t a l Systems Division. 11  Isunevuki M o n t a  Jagan_  1 o k \ o Institute o f 1 c c h n o l o g y  Coordinating Lead  I s u k u b a lbarakL_Japan  \uthor. I A l l WGIII  Chief fiiwiJLTI -ld\ !•*<»• G o \ e m i n e n t o l S n 1 ,mka  Chairman,  V i c e Chair, I P C C Bureau  M u n a s i n g h c Institute f o r  Coordinating Lead  Development ( M I N D ) C o l u m b o . 12 M o h a n M u n a s i n i > h e  Sri L a n k a  \uthor. T A R W G I I I and  SBIM^BI^^^B  Sri L a n k a  Sciretanof  the IPCC,  Intergovernmental Panel o n C l i m a t e Change ( I P C C ) Secretariat I P C C Secretary, B u r e a u 11 N a i a s i m l u n S u n d . i r a r a n u i n  US  Lead Author. S Y R  Geneva. Switzerland  Research Fellow Consultant Division o f Water. and forest  Hmiomnent  lechnologv.  C o u n c i l f o i S c i e n t i f i c a n d I n d u s t r i a l l C" o o r d i n a t i n g L e a d 14. R o b e i l S e h o l e s  South At'nca  R e s c a u h it M K i  Author. I A R W G I and  Pietona. South A l i i c a  SYR  C hid Climate Scwntiir Meteorological Service o f Canada 15 J o h n S t o c k  Canada  rn\ironment Canada  1 cad Author. I'AR W G I  Ottawa. Canada  Review Editor. S Y R  Manager. L u i o p o a n 1 o p i c C e n t e r foi \ i r a n d Climate C h a n g e d  1C A C C ) .  Luiopean I"n\ironmeni Agencv Head o f ! S U I ' A R  16. R o b S w a n  Netherlands  N a t i o n a l Institute o f P u b l i c H e a l t h  WGIII  a n d the 1 n \ n o n m e n t ( F i \ M i  Coordinating Lead  B i l t h o \ e n , T h e Netherlands  \uthor. S Y R  Reseauh Siwrtitt. Global limronmental  Affairs  Department.  \  n  Kanako  Ianaka  Japan  G l o b a l Industrial a n d S o c i a l  Coordinating Lead  P i o a r e s s R e s e a r c h Institute  Author. T A R WGIII  1 okyo. Japan  Lead Author S Y R  National Formal  Interviewees  Representation  Position  WG  Status  Scientist Head of C lunate Anahsii Section. N a t i o n a l C e n t e r foi A t m o s p h e r i c Research ( N C A R ) IS. K e v i n 1 renborlh  Boulder C O , U S A  US  Lead Author. IA R W G I  C h i e f seicnlisi and Director  19 K o b e t i W a t s o n  us;;>. '  l-.nv i r o n m e n l a l l v a n d S o c i a l l y  I P C C C hairman,  Sustainable Development  Cootdinatmg  Network  Bureau  Lead  I he W ' o i l d B a n k G r o u p  Author. 1 A R W G I I and  Washinjilon.DC  SYR  USA.  D i r e c t o i o l the C e n t e r for I-.nergy, h n v i r o n m e n t and L n g i n e e r i n g ( / ) Lead Author. I A R 20  I raneis Y a m b a Government  Zambia  WGIII  Lusaka. Zambia  Officials C hief Idusoi G l o b a l Change Office o l P o l i c y and Iniei'i uinii.ii M L u i U S Department o f F n e i g y  21  RickBradle>  US  '  1 \ R W l . I I I a n d <^ U  Washington D C . U S A ( onsulrunt.  1A R WGIII President o f M a i g a i e e Consultants 2 2 . 1 rie H a i i e s  I cad Author. 1 A R W G I I (Ch  Canada  fi)  loionto, O N Head ol Sciem e-polic v. Global '\tmosphere Division D e p a r t m e n t for 1 n v i r o n m c n t . 23. D a v i d Warrilovv  UK  Food  and R u r a l A f f a i i s  IAR WGIII. S Y R  london, U K  (I A R W G I . W G I I )  Observer Dnetioi. Greenpeace International P o l i c y 24  Bill Hare  International  l A m s t e i d a m . I he N e t h e r l a n d s  LNCiO  Observer  4.5.3 Documents Documents provide a critical, but supplementary source of information in case studies and come from multiple sources. They are located in communiques, memos, agendas, minutes of meetings, administrative documents - such as proposals and progress reports, previous studies or evaluations of the study 'site' and articles in the mass media. These documents can contribute considerable insight to research design and questions. For instance, documents provide clues into important areas of analysis by  64  providing specific details that either corroborate or contradict information from other sources. In the latter, the researcher is provided with explicit reasons to inquire further (Thomas 2003, 80). For instance my interest in this research began with my intrigue over the written mandate of the IPCC taken from the IPCC homepage,  The role of the IPCC is to assess on a comprehensive, objective, open and transparent basis the scientific, technical and socio-economic information relevantto understanding the scientific basis of risk of human-induced climate change, its potential impacts and options for adaptation and mitigation (www.ipcc.ch; emphasis added). The seemingly paradoxical effort to produce objective, neutral and yet policy relevant scientific information intrigued me to explore the process that gave rise to such multi-faceted and potentially instrumental information.  Documents are both the means and result of communications between actors attempting to achieve some common objective. Therefore documents cannot be read as if they contain the unmitigated truth; instead the researcher must always be wary and critical in their readings in order to understand the nuances, unstated assumptions, and local meanings hidden in the document (Northey. et al. 2002, 92). Documents can rarely be used as sources of evidence but instead provide materials from which inferences can be made about the characterizations, activities and networks within a particular research context (Miles and Huberman 1994). Documents in this research are used to corroborate first-hand understandings and descriptions derived from participant observations and interviews.  The use of documents in this research  In this research both formal and informal IPCC documentation is used. It is used in four ways. The first two ways are dependent on formal documents: 1) to determine the official mandate of the IPCC, and 2) to garner an understanding of the formal written processes and protocols involved in the SPM and SYR processes. The remaining two uses involve the informal documentation that includes the discursive interaction that contributes to the final document (see Appendix B): 3) to trace desired changes by each participating expert, delegation and non-governmental organization involved in the review process and 4) to perform content analysis on the iterations of documents using 'track changes' in order to identify the patterns of controversy and resolution in the negotiation of the SPM and SYR texts (see Box 4.2). The example is hypothetical because these intermediary texts cannot be cited or quoted. Although these documents provide considerable insight, the inability to demonstrate these insights in a transparent way pose a slight problem. 65  Box 4.2 A Hypothetical Example of IPCC Inter-Textual Documents  •  The following hypothetical text shows how text might be changed, using "track changes", as part of the discussions in the IPCC Plenaries or contact groups. C L A participants must certify that the final text is in accordance with the text in the underlying W G reports. The underlying text is not subject to such negotiation and line-by-line approval, but is instead 'accepted' as a whole by the IPCC Plenary. The inertia in climate and ecological systems suggests that temperatures will increase will occur in the 21 century. The SRES scenarios^ accompanied by considerable assumptions and, uncertainties, suggest project that in order to stabilize avoid exceeding atmospheric concentrations at levels of 450, 650 and 1000 ppmv, immediate-adaptive strategies and strategies for the mitigation abatement of greenhouse gas emissions will be required. st  Given that the IPCC is intergovernmental, governments are able to review the material, and in the case of the IPCC approve the material as well. There are two formal sets of documents that are accessible for analysis based on these two processes; the review documents and the approval documents (including the final approved SPM (2001)). The review document results from the governmental review process and shows different government comments for each line of the SPM document. This an interesting document that enables the researcher to examine the different comments and desired changes of various experts and governments. In each comment governments are obliged to provide a reason or rationale for the desired changes that must relate to one of three rationales: 1) inconsistency of the information with the underlying reports, 2) imbalance in the representation, or 3) to ensure final messages are not policy prescriptive. These documents provide different rationales about what is considered to be inconsistent, unbalanced or policy prescriptive.  The approval document shows the changes that emerge during the review process and has "track changes" integrated from the changes made in the Plenary session. This document therefore provides a unique method to trace the approvals process. Previous sentences that are tracked out are still readable and added sentences are underlined and thus identifiable. These documents are particularly salient to my research into the SPM and SYR as they each identify areas of change, transgression and thus negotiation. The T A R WGIII SPM and the SYR SPM documents in particular, contribute to the research by revealing the iterations where emphasis and/or omission of material is traceable. The iterations captured by 'track changes' highlight revisions, and make them accessible for analysis. By observing patterns involved in iterations of a document it is possible to explore questions about communications and networking within an organization (Thomas 2003, 80). Although the approval document provides an inter-subjective taste of what was occurring and under review in the Plenary, it does not identify nations involved in the changes nor express in any way the concerns that generated the changes.  66  Although track changes were used on-screen in contact groups, these were not generally made available in hard copy. Therefore these documents are mainly used to corroborate and thus, validate the participant observations and to a lesser extent the interviews. These inter-subjective texts are unfortunately not available for citation or reference, however they provide interesting and relatively novel sources for getting at areas considered controversial and/or policy relevant. They provide resources from which to examine patterns in the process while also providing additional ways to corroborate evidence.  There is a fifth possible area for document analysis that was not accessible or feasible for this project. The rationales used to justify the inclusion or exclusion of expert and government comments were documented and collated by the technical support units (TSU). Content analysis of these documents and the rationales used to make changes would make an interesting research project in itself but unfortunately extends beyond the scope of this study. These scientific rationales could offer insight into the ways that IPCC scientists demarcate science from non-science in the IPCC. The rationales considered important to this research were recorded in my participant observation notes.  4.6 Data Reduction and Analysis There may be a correspondence between life as lived, life as experienced, and life as told, but the anthropologist should never assume the correspondence, or fail to make the distinction (quoted in Miles and Huberman 1994, 267).  The case study researcher must be wary that correspondence of claims using one source of data and/or method for analysis may coincidentally corroborate researchers descriptions. It is necessary to ensure corroboration from multiple sources at a level of pattern and redundancy in order to develop a reliable description of the interactions, practices and process. As an outside observer, the researcher always runs the risk of being ignorant to the particular language, gesture, strategies, meanings, etc. of the culture(s) under investigation (McKinney 2000). Using multiple sources, such as participant observation, interviews, and document analysis, to check findings contributes to a reflexive and thereby more reliable interpretation of the data collected.  4.6.1 Steps involved in reducing the data The participant observations and interviews were transcribed and organized in Atlas hermeneutic software. This software is designed to organize documents and to make coding and access to codes in  67  those documents easier than if hand-done. Both sets of materials underwent two iterations of coding. First, the participant observation and interview documents were reduced through a coding process.  The initial round of coding was done through a deductive process where 12 codes were chosen with which to peruse the data. However I found it difficult to convey the richness of the data with these limited codes and soon found myself developing upwards of 25 more codes. Inductive and exploratory techniques more similar to grounded theory were used to "let the data speak" (Tansey 2004) in this initial coding process. Therefore, once the codes of interest and relevance were prioritized, a second round of coding was completed. These codes related directly to the research interests and the theoretical framework and were thus more useful.  The result of this process was that much of the data collection and reduction process was completed through iterative and inductive means. The judgments made along the way - which scientists to interview, which questions to ask, which issues or controversies to follow, and what codes emerged from the second round of coding - are all illustrative of an unintended grounded theory approach to the research (Tansey 2004).  Multiple layers of interpretation occur in the process of data collection, data reduction and data interpretation. Researcher judgments are made at every stage. This is not to say that the scientific inquiry is dubious. Instead the strength of the interpretive social sciences is to locate the researcher and her particular assumptions that both frame and contextualize the propositions of the research (Risjord 2000). This approach imposes an additional burden of clarity and explicitness of assumptions to the best of one's ability. Indeed, my particular collation, analysis and interpretation of information represent one of many possible interpretations. The assumptions in the research design reflect the commitments held by me, the researcher. Social scientists, similar to natural scientists, are not dealing with static snapshots of the world. Recognizing data as dynamic suggests that it is possible that this research project is located in a particular and possibly ephemeral point in time and particular socio-political context and may thereby become quickly outdated with the need for reflexivity and revision.  39  The overall structure of my dissertation and its particularities, have undergone iterative change since 2001. The initial research questions were exploratory and gave me an overall understanding of the process I was observing and the interpretations of that process. The bounds of the research - the  This is especially significant as the fourth assessment report (4AR) is negotiated and underway. However the character of this research is built around a set of institutions, structures, practices and conventions, many of which are likely to be reproduced as long as the IPCC continues. 3 9  68  questions, sub-questions - have become more focussed since those initial interviews. Time and financial constraints hindered my ability to perform second order, more focussed interviews. Instead I have had to mine the breadth of material and perspectives divulged from the first order (or exploratory) questions and to carefully draw interpretations from the patterns of data that result. Although this hasn't appeared to be problematic, it makes it especially important to have valid (corroboration with other materials) and systematic ways (coding) to ensure that specific interpretations and understandings are warranted.  4.6.2 The interpretation of data  The information in this particular research set is used to support theoretical propositions. It is organized to provide "well-grounded, rich descriptions and explanations of processes in identifiable local contexts" (Miles and Huberman 1994, 1). Narrative format is used to present the original dialogue in the event or interview in order to give the reader a sense of the data itself. "Words especially organized into incidents or stories, have a concrete, vivid and meaningful flavour that often process far more convincing to a reader - another researcher, a policymaker, a practitioner - than pages of summarized numbers" (Miles and Huberman 1994, 1). By finding narratives or storylines, the complex and "sticky" areas of intention and action are revealed in a more enriched and valid way than otherwise defined by particular categories and/or quantifications (Lincoln and Guba 1985).  In order to convey the results in a richer way, I have decided to include narrative text in my descriptions and interpretations of the process. Including the raw data or 'words' as evidence that build on my theoretical understandings and propositions may also increase the reliability and validity of my particular interpretations of the data. It invites the reader to analyze and assess the suitability of the data to my particular theoretical propositions. The collection and interpretation of the data is comprised of particular judgments that have been made at every phase of the research process (see Box 4.2). Descriptions and interpretations based on this data-set may be refuted and/or interpreted differently. If this is the case, alternate interpretations will contribute to a reflexive, iterative, dialectical and I believe enhanced approach to the research process.  69  Table 4.2: The Advantages and Disadvantages of Methods Used  Analytical  Techniques Used  Advantages/ Problems  Components Data Collection  1) Participant  2)  Observation  differences between theory and  Semi-Structured  practice in the process.  Interviews 3)  1) Provided insight into the  2)  Teased out the different meanings, interpretations and intentions of the  Document Analysis  participants (often conflicting). 3)  MS Word "track changes" explicitly reveal the areas where controversy and revision took place.  Data Reduction  1) Initially there were too many  1)  Coding  2)  Identify patterns  codes. They needed to be  3)  Re-code  reworked in more focussed and  4)  Tease out themes  systematic way relating back to  5)  Aggregate  original research questions (2, 3).  6)  Cross-examine with  3) Contradictory and illegitimate  other data forms  patterns were initially given special emphasis - recognize need for balance. 4) Re-code with balance and research focus. 5) Tried to populate, patterns with alternate evidence.  Data Display  1) Code families 2)  Tables and Figures  3)  Appendices  1) Codes helped to identify patterns and to compare information. 2)  Where possible tables and figures are used to give visual representation of the processes and actors.  3)  Appendices provide insight into the kinds of information gathered  and used.  Conclusion  1) Identify areas where  drawing/  data answers research  verifying  questions (also be  2)  Limitations of  2)  Specific evidence is used to  it does not apply).  speculate on why there may be  Identify and account  varied interpretations. (Second  for varied  order interviews were not done as a  interpretations of  source of validity for analytic  participants.  claims.) 1) Prevented the opportunity to hone  interviews were not  more directly onto issues of import  done.  and interviewees to have the  M y lens as  opportunity to comment on  participant observer. I  interpretations. 2)  This was a subjective filter that did  what I considered to  not sequentially deal with data  be important to the  reduction techniques.  research questions.  4)  2)  corroborate analytic claims and to  only marked down  3)  theoretical propositions.  wary of areas where  1) Multiple iterations of  Research  1) Patterns generated contribute to  3)  Representation would enhance  Lack ofWGII  exploration of possible differences  participants.  between W G participants'  Chose which contact  impressions.  groups and  4)  A d hoc attendance influenced my  discussions to attend  experience and observations in the  in an ad-hoc way.  process.  71  4.7 Implications of this Analysis In this chapter the rationale for a constructivist approach was explained as a way to examine differences between what is idealized in theory and what occurs in practice. In the SSS, a constructivist approach is used to investigate differences between theories of scientific knowledge and science as practice. The focus on science as it is practiced identifies social, conceptual and material processes that would otherwise be omitted from or purified from the final scientific interpretation. This approach is advantageous when using a case study methodology. It enables the researcher to make distinctions between what is formally represented through final claims and representations and the informal processes and practices that influence the interpretation and representation of those claims. Case study techniques provide opportunities to collect data on the historical, inter-textual and informal processes and practices involved in case study analysis. A descriptive analysis is in general required for a prescriptive analysis to occur. The techniques used to collect this type of data include: participant observation, participant interviews and document analysis. These three data collection techniques are used in the examination of the IPCC process in order to decipher between what is said about the process and what is unsaid and omitted from the final representations. The IPCC was chosen as a case study due to its focus on process in crafting the final representations of the policy relevant and policy neutral SPM and SYR documents.  4.8 The Following Chapter Historical analysis is particularly important when using case study methodology (Yin 1984). It is necessary to understand how the particular setting that is under examination was initiated and how it formed. For this reason the following chapter will explore the history of the IPCC and the formal developments and adaptations the process itself has taken over its past three iterations (1990, 1996, 2001).  72  CHAPTER V: Conventional Science-Policy? Understanding the Governmental Presence in the IPCC This Chapter will examine the ways that the Intergovernmental Panel on Climate Change (IPCC) has evolved over the past three iterations; the first assessment report (FAR), the second assessment (SAR) and the third assessment report (TAR). The ways the IPCC has generated credibility and legitimacy for its products is directly related to the ways it has adapted the assessment institution to criticism from external forces. By being resilient and adaptive the IPCC has derived its credibility and legitimacy based on inclusion rather than exclusion of conflicting viewpoints and by including an increasingly diverse user community. How has the IPCC enrolled scientific and political allies into the scientific assessment process while maintaining the credibility and legitimacy of the products?  The IPCC was created in 1988 as an intergovernmental scientific assessment that has, through various adaptations in the process, evolved into a body made up of its own practices and form of inquiry. Procedural management by the IPCC's Bureau attempts to simultaneously ensure the efficacy of the 40  IPCC, the credibility of the scientific information produced and the legitimacy and relevance of the information to governments. To accomplish these objectives the IPCC has had to be adaptable not only to emerging and rapidly increasing climate science literature but also to the diverse needs of the international policy sphere. What is interesting is that in maintaining distinct and formal boundaries between science and policy communities, the IPCC has managed to increase the governmental presence and processes for science-policy interaction throughout its structures. This chapter explores the ways the IPCC has reflexively adapted, modified and evolved its formal procedures to more effectively mediate between the worlds of science and policy. It is interesting to note that many of the adaptations that have occurred have been ad hoc responses to contextual pressures (see Table 5.1). The end result has been a process that, over its three iterations, has increasingly included governments within its structure and processes in efforts to produce "policy relevant but not policy prescriptive" information. The following history of the IPCC reveals that in areas of scientific uncertainty and controversy, a particular scientific interpretation becomes stabilized via 1) alliance-building both internal and external to science and by 2) developing strong procedures for addressing dissenting scientific views. The climate assessment has evolved to include the international policy community more and more actively over its three iterations. In the T A R (2001) and T A R Synthesis Report (SYPv) (2001), policymakers were formally embedded into the process through formal procedures and guidelines (IPCC 1998) to the extent that "the products of the  The Bureau is made up of the Chairman, the Secretary, six Co-Chairs (one developed nation co-chair and one developing nation co-chair for each working group), and seven elected members.  4 0  73  IPCC are not official documents until they are approved by the international policy community". The salience, credibility and legitimacy can be increasingly stabilized if negotiated and crafted by the 'user' community in effective ways (Jasanoff 1990, 1991, 1996; Clark et al. 1997; Cash 2001).  Table 5.1: Differences in Structure Between the First Assessment Report (FAR) and the Second Assessment Report (SAR) FAR  SAR  Informal, close/ science-policy interaction  Formal science-policy interaction  Small epistemic network  Large, diverse international participation  Focus on urgency; policy action now  Focus on uncertainty; wait and see approach  Scientistic framing  Integration of social values (historical responsibility, costs and benefits)  With the increasing inclusion of the knowledge users in the process has come other responsibilities for the IPCC Bureau, such as 1) managing the autonomy and impartiality of the scientists in order to secure scientific integrity and scientists' involvement, 2) formalizing the boundaries between scientists and policymakers, and 3) internalizing external dissent and conflicts in the production of policy relevant scientific information. In order to explore and examine the ways the IPCC has responded to these responsibilities a bricolage of methods is used to reconstruct the assessment's procedural history. This includes a combination of existing analytical literature on the IPCC, formal and informal IPCC documents and interviews with veteran IPCC participants (scientists and Bureau members).  5.1 Including the User: Enrolling the Policy Network  From the intergovernmental mechanism that situated the climate assessment in an international governmental framework to the derivation of the policy relevant scientific questions (PRSQ) that frame the S Y R of the T A R , the IPCC has from its inception developed processes for interfacing with policy. Although this is not unusual in mandated science processes such as assessment and advisory bodies that collate information and make interpretations for a user community, the IPCC has gone a step further by introducing international governments into the scientific assessment process itself through the Summary for Policymakers (SPM) and S Y R processes (IPCC 1998, 1999). On first glance this may be viewed as a contamination of the science, however on closer examination a more nuanced crafting of information is taking place that aims at ensuring that the IPCC products remain credible to the scientific community,  74  legitimate to the international policy community and to ensure that the climate information remains salient to the international policy environment. In this way the IPCC not only interfaces with policy in the conventional ways of an assessment institution but also straddles and manages the precarious nexus between science and policy or experts and governments.  The history and development of the IPCC reveals the ways that the assessment of climate information is dependent on the existing socio-political sphere for which it is being assessed. The IPCC was developed as a science-policy interface, where the scientific consensus taken at the 1985 Villach conference could be translated and made relevant to the international policy community (Agrawala 1999a,b; van der Sluijs et al. 1998). The desire to "appeal to politicians for action" (Bolin 1994) initially generated an epistemic forum where the major industrial nations were targeted for support. However, the initial activities of the IPCC were not without controversy. In particular, the exclusion of national representation from the . majority of nations (developing nations), the scientific controversy over methods, instruments and interpretations that arose in the media, social and policy analysts' concerns about politicization and the close working relationship between scientists and governments gave rise to a number of criticisms which in turn led to several procedural innovations. These innovations increased the participation of governments in the assessment process while at the same time formally managing the boundaries and roles ascribed to scientists and governmental delegates within the process.  This chapter explores the trade-offs involved in fostering scientific credibility simultaneously with political legitimacy. Although both are required in order to mobilize scientific information in the broader socio-political community, very little empirical analysis has been undertaken that examines processes for achieving this at the international scale.  41  It is clear that only if information is seen to be legitimate by  knowledge users can scientific information provide an authoritative foundation to contribute to meaningful discourse and cooperation among governments in the international sphere. However in efforts to remain legitimate to knowledge users, how are the practices of knowledge producers altered?  5.2 History of the Science-Policy Interface Climate change was initially a scientific issue examined by meteorologists and atmospheric scientists interested in determining the fluctuating greenhouse concentrations in the atmosphere. It was an invisible problem, both to the unspecialized eye and to the political world. This changed in 1982 when the Advisory Group on Greenhouse Gases (AGGG) was formed by the International Meteorological Institute in Stockholm to review the state of the considerable body of climate science that had been developed both Farrell et al. (2001) is one of the few such analyses. He examines the different processes and effectiveness of six multilateral environmental assessment processes including the IPCC.  4 1  75  at national and international research institutions in order to define patterns and to consider the possibilities of anthropogenically-induced climate change (for specific understanding of the history of climate science and the creation of the IPCC see Boehmer-Christiansen 1994a,b; Agrawala 1998a,b). The activities of the A G G G culminated in the 1985 international 'Conference on the Assessment of the Role of Carbon Dioxide and of Other Greenhouse Gases in Climate Variations and Associated Impacts' in Villach, Austria (1985) sponsored by UNEP, the W M O and the International Council of Scientific Unions (ICSU), which brought scientists from all over the world to form an international panel on climate change (van der Sluijs et al. 1998).  New scientific information was presented at this conference that illustrated the role that other greenhouse gases (GHG) beyond carbon dioxide play in enhancing the greenhouse effect (Ramanathan et al. 1985). With this additional information, scientists anticipated that the radiative forcing, corresponding to a doubling of the C 0 2 concentration, of the atmosphere would occur by about 2030 (Bolin 1994). This accentuated the urgency of the climate situation and mobilized scientific knowledge brokers to present findings in the international policy arena. This conference and international consensus is credited with having initiated the climate regime (Bolin 1994, Agrawala 1998a).  The international scientific consensus of the Villach conference was actively used to generate normative concern and political will. Analysts dispute whether the consensus was forged based on the contribution of the Ramanathan et al. (1985) paper or alternatively whether it was a collective environmental ideology among an epistemic network of high profile scientists that lead to the scientific consensus (Agrawala 1998a; Boehmer-Christiansen 1994a,b, Miller 1997).  42  Either way the authority demonstrated through  international scientific consensus and the active efforts among reputable international scientists to use existing networks of influence resulted in successful efforts to construct a policy environment in which to translate scientific information (Agrawala 1999a). The development of an epistemic coalition was facilitated by the Villach consensus where a common interpretation of the science was conveyed to powerful politicians and bureaucrats (e.g.. former British Prime Minister Margaret Thatcher). This created a shared understanding of the problem and common values and policies for its resolution (Haas 1992). Whether or not this consensus was due to the addition of scientific information or to the epistemic commitments to extend the findings into the international policy realm is unclear. However the original IPCC Chairman Bolin (1994, 26) comments that, "the outcome of the [Villach] conference was an appeal to politicians for action" (1994, 26). Subsequently it has been argued that this conference was a "major step in interfacing science with policy" (van der Sluijs et al. 1998, 292). Miller (2004) argues that the threat of climate change had more to do with the environmental ideology of the scientists, who actively constructed the climate problem in the international political sphere, than it had to do with the extension and development of the state-of-the-science.  4 2  76  Former US Federal Aviation Agency (FAA) Director and long-time IPCC Secretary, V . Sundararaman, wrote the initial resolution to the W M O and UNEP, requesting sponsorship for a comprehensive assessment of climate change.  43  The comprehensive assessment was considered to be a way to link  climate science in a more relevant way to the values and interests of policymakers. Through an expanded working group (WG) structure, a knowledge continuum is generated that moves information from the natural scientific realm through to the applied and social scientific realms (and policy-related realms). Climate science (WGI) is used to understand the impacts (WGII) and socio-economic responses (WGIII) to climate change, thus ensuring the practical relevance to the international policy community. In 1989, the resolution was first passed by the W M O and two weeks later by the U N E P for the development of an international climate assessment. This process was not entirely seamless however.  5.2.1 Development of the Intergovernmental Mechanism The W M O was considered to be the appropriate international institution to house the assessment in order to de-politicize the scientific assessment. UNEP was seen as a necessary sponsor in order to coordinate the international policy audience. However UNEP was struggling with its scientific and political legitimacy due the perceived ideological directives of the executive director, Mostafa Tolba in the ozone assessment (Agrawala 1998a,b, Interviews [38], 2001).  44  Due to what was perceived as UNEP's political  and/or ideological interest in the climate issue, there was difficulty in getting some nations, particularly 45  the US, the biggest financial patron to UNEP and the biggest greenhouse gas emitter, on side with an international climate assessment (Agrawala 1998a). The only way to bridge the distrust and skepticism that had developed between the US (among others) and U N E P in the ozone assessment was to ensure a transparent climate assessment process that included governments in the review of the scientific information. Recollecting this process, a veteran participant recalls,  The National Academy of Sciences (NAS) had used a three-pronged approach to assessment examining not only climate science but also potential impacts and response options. IPCC Secretary, Sundararaman and others believed this was a good way to contribute to the political relevance of the information (personal communication, March 4, 2001). The literature on the historical developments of the IPCC is corroborated by interviews with veteran IPCC participants and Bureau members. Interviewees' names are withheld in order to meet U B C Ethical Review requirements. The annotation used for quotes includes a bracketed number that symbolizes the identity of a particular author and the year the interview was performed. The bracketed number is used mainly for the ease of the researcher, to keep comments organized, but can also be used by the reader to view the range of authors and their diverse perceptions and/or to trace the comments of a particular author. Once the ozone assessment was completed, U N E P had already formulated an agenda of future issues with climate change at the top.  4 3  4 4  4 5  77  [The US] actually said... that they needed a government mechanism except they used the word ad-hoc because they didn't know how long it should be or how long it should last and various questions and what exactly should be the terms of reference (Interview [38], 2001). The requirement of an intergovernmental process satisfied sceptical governments' needs to be involved in reviewing the information produced.  46  By including governmental representatives in the scientific  assessment process, two objectives were fulfilled: i) scientists coordinated or enrolled a governmental audience to which to transfer their climate findings and ii) participating national governments had the ability to review and authorize the legitimacy of the process; granting governments access to the arena of knowledge production. The ability to review and evaluate knowledge created a space for the maintenance of sovereign authority in the process; each nation was able to review and comment on the information presented. This was necessary in order to develop legitimacy in the scientific process and to increase the possibility for stabilizing knowledge in the international policy sphere.  47  More importantly the derivation  of the intergovernmental mechanism established a working relationship between scientists and the 'big players' (thus the 'big emitters') in the international policy community.  Through ad hoc, and what was then considered to be short-term measures, the US, USSR and U K were delegated the responsibility for coordinating and providing technical and financial support for each of the respective working groups (WG I, II, III). Governments were not only introduced to informally review information but were fundamentally relied upon for their financial commitment. The financial dependence of the IPCC on individual governments has continued throughout its three iterations. This is done through the establishment of technical support units (TSU) where governments oversee the managerial and bureaucratic functioning of the construction of the IPCC reports. This delegation of responsibility has fulfilled two major functions: i) it has enabled the IPCC Secretariat to remain small, operating on less than 650,000 SF or $350, 000 U S D per year, where a portion of this money is used to support developing nation participation, and ii) it helps to prevent the IPCC from falling into the trap of many organizations that become institutionalized and continue to support their bureaucracies despite completing their initial objectives. After the completion of each assessment the decision as to whether the IPCC should continue remains in the hands and pockets of governments and is therefore dependent on the relevance of the reports to governments. One WGUI author notes that the success of the IPCC is determined by "the fact the governments continue to support it and therefore [they] must find it useful"  Without knowing how to develop an intergovernmental process, the WMO Secretary General, Obassi, invited governments participating in either the WMO or UNEP to discuss climate change. The first IPCC meeting was held in May 1988 where an international community of scientists was brought together and where 32-33 delegations attended "to collect technical information relating to climate change" (Interview [38], 2001). The timeline for the completion of FAR was intended to allow completion in time for the Second World Climate Conference (1990) (the first was held in 1979). Cognitive authority is achieved through "access to knowledge and the ability to evaluate it" (Nowotny 1992).  4 6  4 7  - 78  (Interview [42], 2001). This a critical point when discussing the relevance and legitimacy of climate science (referred to in more detail later).  The coordination of scientific reports for considerable political events has been a strategy in the development and promotion of IPCC reports. In the development of the IPCC, scientists actively prepared the first series of Climate Change Reports for the World Climate Conference in 1990. The Response Strategies Working Group (WGIII) (now referred to as Mitigation) assessed and interpreted information that could be used for the development of a Framework Convention on Climate Change. The intergovernmental negotiating committee (INC) formed in order to initiate the development of this framework before the United Nations Conference on Environment and Development in Rio de Janeiro in 1992. Since then, the IPCC has synchronized its reports with activities of the U N F C C C and COP activities. The parallel scientific and political processes have contributed to the long-lasting issue attention cycles in the climate issue.  48  It can be argued that the integration of governments in the review of the products and in the fiscal responsibility for these products, combined with the strategic timeline, has contributed to the saliency of the IPCC assessment in the international policy sphere. Each of these factors has contributed to the development of a considerable interface between the scientific and governmental communities. When asked whether the IPCC was conceived of as an experiment in developing a science-policy interface, the IPCC Secretary asserts that,  No, the idea always was from the beginning that the assessments should help policy formation but we shouldn't do the policy formulation. We should stay away from that. No the feeling was very strong about the separation of church and state so to speak ... There should be a fine-line between the assessment - the fact-finding activity and the policymaking activity. You cannot taint the former with the latter, people suspect at that point (Interview, 2001). Despite this strong indication that science is separate from policy, the initial IPCC report, the first assessment report (FAR), came under considerable criticism. Charges of a close working relationship between scientists and policymakers in the F A R (O'Brien 1990; Next Steps 1990; SEPP 1991; BoehmerChristiansen 1994a,b) were made after its publication. This was especially true of WGIII that was seen to be populated by mostly civil servants (Aldhous 1990) raising conflict in the F A R (and later in the SAR) around the credibility of the scientific representation (Masood 1997). Hilgartner and Bosk (1988) identify that problems in the social sphere are collective sentiments not mirrors of objective conditions and they require consistent and collective attention to be considered a legitimate public/policy problem. In the public arena problems compete for public attention and resources. Down (1972) discusses issueattention cycles in the public sphere. Hefindsthat once support for a particular environmental policy issue peaks, there is very little consistent or long-term attention paid thereafter, as other issues override its importance. 4 8  79  The combination of limited participation in the assessment and the robustness and urgency of the conclusions drafted, caused concern among scientists and policy makers alike over the influence of an ideological and environmentally-driven epistemic community (Aldhous 1990; Boehmer-Christiansen 1994a,b). For example an anonymous opinion piece was published in Nature stating that,  The goals [of the IPCC] should not be the persuasion of all politicians to a simple common view of what the threat of global warming entails, but to a rounded understanding of the complexities of the problem with which the world is faced, and to an understanding of how it may be ameliorated (Next Steps 1990, 182). These initial criticisms had the effect of reducing the credibility and legitimacy of the substantive reports in some international circles. Developing nations attacked the process as another form of northern elitism (due to their lack of involvement) that promoted a colonial-environmental agenda (Kandlikar and Sagar 1997, 1999; Lahsen 2001; Biermann 2001); and external critics questioned the procedural legitimacy that enabled scientists and policymakers to simultaneously construct and review climate reports (BoehmerChristiansen 1994a,b). A n IPCC WGI scientist, Ralph Cicerone, cautioned against high expectations of the initial IPCC reports. He was quoted as saying that expectations from a "hasty study like IPCC may have been too high", and that the "the IPCC is the first step not the final answer for international agreement" (Aldhous 1990, 373). This interpretation does not suggest a quick or global solution to the climate problem but rather views the assessment as an initial step in ongoing efforts to compile and assess information on climate change.  The underlying evidence that the climate was changing due to human interference was brought into question at this time primarily by US climate scientists.  49  For instance, US climate scientists (namely  Fred Singer) sceptical of the climate science issued by the IPCC and its circulation in the political process, in turn, released the Heidelberg Declaration at the United Nations Conference on Environment and Development (UNCED) Rio Summit (1992). The Declaration expressed "a conviction that modern society is the best equipped in human history to solve the world's ills, provided that they do not sacrifice science, intellectual honesty, and common sense to political opportunism and irrational fears" (SEPP 1992).  50  Over two hundred scientists and experts, eliciting further and increasingly public challenges  Climate sceptics such as Fred Singer, Richard Lindzen and Patrick Michaels have written op-ed pieces, provided testimonials to the US government and have circulated caution over taking climate science too seriously. The arguments lie in the inability to validate the climate models and in the emphasis of these models in IPCC reports, including the SPM. The concern is that inadequate models are being used to influence international policymakers to take premature and potentially costly policy measures. Also on the Science and Environmental Policy Project (SEPP) website (http://www.sepp.org) are various articles, either published in newspapers or not published at all, and exchanges between climate sceptics and IPCC  4 9  5 0  80  against the credibility of the underlying climate science, signed it. The criticisms that mounted after the publication of the F A R reveal sceptical attitudes toward the underlying scientific premise of climate change and notably for the purposes here, the way that science was interpreted, communicated, and diffused into the policy sphere.  On one end of the spectrum were those that challenged the content of the science, highlighting the considerable uncertainties in climate science and arguing that it is therefore insufficient for the policy sphere. This view suggests a positivist approach to science-policy whereby science must be objective, robust and consensual before being transferred to the policy sphere (Victor 1990; Singer 1995). hi his presentation to the U S House Committee on Science, Space and Technology, climate sceptic Richard Lindzen (1991) argues against the scientific efficacy of the IPCC WGI report. He also notes that the preparation of the Supplementary Report to the IPCC Scientific Assessment (1992) so close to the original publication (FAR 1990) is indicative that the "original document was hardly canonical" (1991, 1). Much of the critique comes from a concern about the IPCC's status in and alignment to the policy community. In the positivist view, there is a concern that alignment to policy politicizes the content of science and therefore diminishes its objectivity, autonomy and impartiality (or neutrality). This same scientist illustrates this concern in a more recent quote,  Science in the public arena is commonly used as a source of authority with which to bludgeon political opponents and propagandize uninformed citizens. This has been done with both the reports of the IPCC and the NAS. It is reprehensible practice that corrodes our ability to make rational decisions (Lindzen 2001b). The positivist understanding of science advanced here promotes a unidirectional model of science in policy whereby science undirectionally flows into the policy sphere to contribute to rational decisionmaking. On the other end of the spectrum there were those who challenged the authority of IPCC science in the policy sphere. In this view there was "concern about hasty policy action based on inadequate science" (Singer 1995). This was a procedural concern that IPCC climate science was presented in the SPM as if it were robust and/or conclusive, thus giving the reports authority in the international policy community.  51  In this view the scientization of policy was also seen as a threat similar to the way that MacKenzie (1990) describes in his 'certainty trough' where underlying uncertainties can be downplayed and scientific  participants that reveal the flurry of activity that was taking place after the publication of the FAR, and later following the publication of the SAR. This sceptical understanding can emerge from the scientific sphere, where positivist scientists refuse to align within forums of mandated science and believe all scientists that do so, contaminate their value-neutrality and objectivity (Singer 1995; Lindzen 2001). Or it can emerge from the political sphere where the interpretive flexibility of science in the policy sphere is seen to increase rather than decrease complexity in the decision-making process (Jasanoff 1996). 5 1  findings accepted as robust in the policy community. This was of particular concern among developing nations, who saw the 'use of science' for international policy as in thrall to powerful economic interests, or as distorted by those interests through the IPCC process. What is conveyed through these criticisms is that if one takes a positivist view of science and a unidirectional model of science-policy, the motivation of scientists' involvement in mandated science and the information produced can become areas of contestation bringing issues of trust and competency into question more than the issue or problem at hand (Shaw and Robinson 2004). As indicated in Chapter 3 this is the problem with the scientization of policy; it inevitably leads to the politicization of science.  5.2.2 Making Global Reports Relevant  According to a recent global e-forum initiated by The United Nations Development Program (UNDP) from April 6 through to May 10 2004 (2004) on "Global Reports and Global Policy Challenges: Match or Mismatch?" suggested that it is very rare for global reports to be diffused to the policy sphere. The general sentiment among policymakers, academics, students, officials and researchers from over 35 nations, was that most global reports and projects largely go unnoticed. Many speculated that this ineffectiveness was largely due to the voluminous size of global reports that contributed to the lack of accessibility and dissemination. The lack of alignment with the policy community was identified as a problem contributing to the lack of interest, the disconnection between research and the needs of the policy or user community. It was concluded that global reports are informative, can be critical for analyses of the causes, interactions and drivers that contribute to global environmental change, but must be made more accessible and salient to the international policy sphere  Based on this e-forum it became clear that the kind of continual policy concern, research, deliberation, and media attention that the IPCC has garnered over its 17 years of activity is a relative anomaly compared to most global efforts. Part of what has contributed to both its criticism and significance in the international policy sphere is its inclusion of governments through the intergovernmental mechanism. Another aspect contributing to this attention is the IPCC Bureau's willingness to reflexively and iteratively adapt the process in order to internalize criticism namely through the policy relevant innovations that emerged in the SAR. The SPM and SYR were created to ensure that a condensation and synthesis of the key policy relevant findings took place simultaneous to the drafting of the technical summaries and reports. In the written words of the first Chairman, Bert Bolin:  Scientists need to inform politicians in a simple manner that can readily be understood, but the scientific message must always be scientifically exact. In reality, little of what 82  we know as scientists is politically interesting or even understandable. Politicians are seldom scientists. It is difficult to sift objectively all the available scientific information and extract what is politically relevant. For that reason, the IPCC has chosen to publish each of its three working group reports in two parts: i) Supporting material of about 300 pages, which is a careful scientific overview of the available literature with numerous references to original articles. It is to be extensively reviewed by the scientific community to ascertain its credibility. ii) A policymaker's summary of about 25 pages, which is an extraction from the supporting material of what may be politically important. It is similarly extensively reviewed by the scientific community (1994, 26). This rigid distinction between the scientific products and the political products was necessary in order to 1) attract reputable scientists to the IPCC process, 2) to ensure that the scientific and policy communities still recognize the IPCC as a credible scientific assessment process through its autonomous scientific practices, and 3) to ensure that the scientific information is accessible to policymakers through the summary of relevant findings.  5.3 Constructing the Second Assessment Report In the SAR, formal procedures for external and governmental review were instituted. A considerable effort to distinguish between the scientific products of the IPCC and the relevant governmental products generated two distinct types of documents: (i) the scientific documents - the underlying W G reports, and (ii) the policy documents - the SPM for each W G report. The procedures changed to project the impartiality required to maintain the credibility of scientific collations and interpretations and to provide the legitimacy required among policymakers. A Synthesis Report that bridged all three W G reports was also introduced in the SAR to connect information more closely with the parallel United Nations Framework Convention on Climate Change (UNFCCC) multilateral negotiation and more particularly Article 2 that examines critical thresholds and the concept of 'dangerous' anthropogenic interference.  52  This was intended to increase the accessibility and relevance of the underlying W G reports, Special Reports and other IPCC documents.  The changes in the structure are also reflected in changes in the underlying scientific messages from the F A R and SAR. The effort in the F A R to present a unified consensus in order to initiate international concern and cooperation on emission reductions policy, led to undesirable skepticism about the  The ultimate objective of the UNFCCC, as expressed in Article 2 is: ... the stabilization of greenhouse gas concentrations in the atmosphere at a level that would prevent dangerous anthropogenic interference with the climate system. Such a level should be achieved within a timeframe sufficient to allow ecosystems to adapt naturally to climate change, to ensure that food production is not threatenedd and to enable economic development to proceed in a sustainable manner (http://www.unfccc.de). 83  information produced and led to speculation about the modus operandi of the participating authors involved. Much of the concern initially arose out of the lack of reference to underlying uncertainties in the IPCC information communicated to the international policy sphere. This robust interpretation was considered to be a policy prescriptive attempt by scientists to influence international environmental mitigation policy (Singer 1991; Boehmer-Christiansen 1994a,b).  The message of the F A R revealed the sense of urgency, certainty and consensus felt by a core group of scientists. It also provided policymakers with the dramatic and prescriptive message requiring governments to cooperate through the development of a framework convention (developed in the F A R WGIII 1991) in order to take immediate action to reduce greenhouse gas emissions.  We are certain emissions resulting from human activities are substantially increasing the atmospheric concentrations of the greenhouse gases: carbon dioxide, methane, chlorofluoro-carbons (CFC's) and nitrous oxide. These increases will enhance the greenhouse effect, resulting on average in an additional warming of the Earth's surface... The long-lived gases would require immediate reductions in emissions from human activities over 60 percent to stabilize their concentrations at today's levels (McG Tegart et al. 1992, xxv). This policy message is considerably different than that in the SAR. The difference comes not only from the message but also from the underlying process that contributed to the approval of the message. The inclusion of governments in the approval of the SPM and its representation contributed to a more politically legitimate message that included associated uncertainties. This message states that, "humans have had a discernible influence on climate change" (Watson et al. 1995, 26). The ability for governments to interrogate the WGII SPM in Shanghai in 1996 enabled delegations to ascertain that the original representation, 'significant influence', was a misrepresentation of the underlying evidence. Key political officials argued that the term 'significant' represented a very specific scientific concept based on the significance of statistical probabilities. It was duly noted that the underlying evidence was not based on an understanding of probabilities and it was therefore argued that it be changed to reflect the science in an unbiased way.  The term significant was changed to the less authoritative yet more accurate and  reflective term 'discernible' (Interview [44] [41], 2001).  Whereas in the F A R the focus was on 'action now', the message in the SAR focused more closely on the underlying uncertainties in the scientific information leading to or facilitating a wait and see approach to policy negotiations. Boehmer-Christiansen (1994a,b) argues that the inception of the international climate research program (culminating in the IPCC) was motivated by two incentives — the ability to secure funding and the ability to coordinate and promote an environmental policy agenda. In the F A R , she argues, that scientists aligned themselves with an environmental agenda of "action now" supported by 84  a small yet significant group of upper level bureaucrats (Boehmer-Christiansen 1994a). Yet she indicates a change in the SAR wheregreater significance was attributed to scientific uncertainties in the SAR, which promoted a "wait and learn" approach to policy. Boehmer-Christiansen suggests that, "faced with the complexity of environmental science researching at the frontiers of knowledge scientific advocacy can honestly switch from emphasizing certainties to uncertainties, from the advocacy of 'action now' to a 'wait and learn' approach" (1994b, p. 197; emphasis added). She argues that the switch to a 'wait and learn' approach benefited IPCC scientists interested in securing funding for their climate research programs.  53  However Boehmer-Christiansen's analysis does not allow for changes in scientific understanding between the F A R and SAR, or account for the different actors and participants that were included in the SAR (Shackley and Skodvin 1995). With a broader inclusion of scientists and experts, wider ranges of scientific judgments were also included about the adequacy of the methods and approaches and the sufficiency of the interpretation for the policy sphere. Also unaccounted for and more critical to this thesis is that the increasingly formalized process included member governments of the W M O and U N E P in the review of the technical information. The concerns and criticisms of the F A R and the involvement of governments in the process through the intergovernmental mechanism put pressure on the IPCC to more carefully to patrol the boundary between science and politics, thus leading to a greater emphasis on communicating the uncertainties in the science in the SAR. These concerns also highlighted the ways that a conventional unidirectional model that focuses on producing consensus-based science to influence the policy sphere was insufficient to compel governments to action. It raised new questions about a more appropriate model that accounted for both the credibility and the legitimacy of science in the policy sphere (see Shaw and Robinson 2004).  Since the IPCC was constructed to ensure climate change became an issue of priority in the international policy sphere, significant efforts were taken to internalize criticisms of the process in order to increase the legitimacy of its message among a more representative group of international policymakers. The inclusion of governments in the review and SPM process through the intergovernmental mechanism was a way to enable governments to monitor the selection and interpretation process from afar in order to secure the legitimacy of the underlying documents (Interview [38] [42], 2001).  The inclusion of governments also meant that the reports shifted from a scientistic framing of the problem where the critical role of the IPCC was to transfer information into the international policy sphere.  A number of scientists interviewed commented that the funding has increased in this field and that more opportunities have been made available via the IPCC including having work referenced and cited in the underlying assessment reports. 5 3  85  Instead it shifted to more of an integrated and inclusive process where governments with differing policy frameworks, goals and values were involved in negotiating and approving the final representation. Once this formal interface was established through the SPM and SYR, different governmental interpretations of the normative issues of climate change became an embedded part of the process (this will be described in more detail in Chapter 6).  5.3.1 Working Group Report Review Process A n elaborate review process was formalized in the SAR in order to preserve the autonomy of the underlying W G reports from any imbalances in the selection of underlying information. There are three rounds of formalized review procedures in the underlying W G reports. The first draft of the information is collated, synthesized and presented through the work of coordinating lead authors (CLA) in the W G and the logistical services of the T S U . This is referred to as the zero order draft (ZOD) and is circulated to IPCC participants for internal review with respect to the representation of the technical information. At this point lead authors (LA) in individual chapters are asked to produce a new 'first order draft' (FOD) incorporating responses to the review comments and identifying underlying controversies or areas of disagreement in the literature. The protocol states that these areas be highlighted in a transparent way through the use of footnotes in the report. The F O D is then circulated to external experts for scientific review and comments. One scientist laughed as he speculated that the majority of worlds' scientists involved in climate research are either authors in the IPCC or are expert reviewers for the process (Interview [42], 2001). Revisions are made based on the comments received and they are collated for the second order draft (SOD), which is then circulated simultaneously to experts and governments for review and comments. Expert reviewers are enrolled to evaluate the scientific and technical credibility and comprehensiveness of the information and governments are enrolled to evaluate the legitimacy of the information, thereby providing checks and balances in the process. This encourages reviewers, including governmental reviewers, to be well versed in the scientific literature in order to be able to identify missing literature or biased interpretations of the literature and to provide comments for revision. The third order draft (TOD) is then presented in the W G Session and then the Panel Session for acceptance. Given the technical and voluminous nature of this research detailed review is impossible at this stage and "acceptance" of the underlying reports signifies that the content is accepted as a balanced and representative review of the relevant peer-reviewed literature. This external review process created constitutive protocols that provide the basis for perceived scientific autonomy, impartiality and valueneutrality.  86  5.3.2 The SPM: The 'Translation'Process  The IPCC process expanded to include the policy 'user' community in more formal and explicit ways both in the review of the underlying technical information as well as in the review and drafting of the SPM. The SPM, considered the document of the governments and the most concise and broadly used product of the IPCC, requires that governments approve, ratify or 'sign-off on the underlying policy relevant messages. According to Bolin (1994) the SPM was to be the point of transfer from scientists and translation by governments of "politically important" material. The Co-Chairs of the W G have considerable oversight over their respective W G reports, due to their editorial roles in the underlying reports and are therefore responsible for writing the initial draft of the SPM, though they typically ask for summary material from their chapter Writing Teams as input to that process. This draft goes to expert and government reviewers simultaneously. Comments are made on the basis of three criteria to ensure: 1) consistency with the underlying report, 2) a balanced treatment of the science, and 3) a policy neutral message. The review comments come back to the authors, who must integrate those comments that are not demonstrably wrong and provide rationales for the inclusion and exclusion of each comment. This draft is then presented to the IPCC Plenary where it undergoes line-by-line approval by governments using the same three criteria. Each participating nation has veto authority therefore ensuring consensual and unanimous approval in the Session of the Panel. In order to accomplish this, formal UN-style consensus negotiations were set-up to approve the 20-30-page document line-by-line.  Perceived bias in the representation of the science and its message is crosschecked through systematic 'due-process' where concerned delegations negotiate different wording while the relevant scientists sit at the front of the room to ensure the scientific credibility of the changes. This formal structure separated the scientific activities of W G assessments from the key policy relevant findings of the SPM and facilitates a formal negotiation at the interface to ensure the document has formal buy-in from all participating government delegations. This complex process uses a ratification process to contribute to the stabilization of collectively approved documents that are then circulated to policymakers and others concerned with climate change, worldwide (the EPCC is translated into the 5 U N languages other than English).  54  The five UN languages are French, Spanish, Arabic, Chinese, and Russian.  87  5.3.3 The Synthesis Report  The SYR also emerged in the SAR in order to collate the key findings in all three working groups and to convey these findings in a document that attempts to bridge the disciplinary silos and to create a more consistent, comprehensive and cohesive narrative for the international policy sphere. In the SAR, a W M O executive council decision to make more explicit links between the findings of the underlying W G reports and the objectives of the U N F C C C , with specific focus on Article 2 resulted in a Synthesis Report (IPCC, web). The SYR had a similar structure to the underlying reports including a similarsynfhesis of policy relevant findings in the SYR SPM. The difference is that rather than an assessment of existing literature the SYR is an assessment of the underlying assessment reports.  5.4 Constructing the Third Assessment Report (TAR) In the T A R a clear and formal set of procedures was created to ensure standardized practices among participants within the institution. It can be speculated that these IPCC processes and procedures were developed in part as a response to charges of violations of due process in the SAR.  Sceptical forces  publicly accused C L A Ben Santer, of changing information after it had already been approved through the intergovernmental process and before it went to print (Interviews [38] [45]). This was highlighted as a breach of confidence in the IPCC process and in the legitimacy of the summary information (see SEPP website http://www.sepp.org). Although there are declarations made that support Santer in making editorial changes, the public extent of this attack reflected poorly on the process (Edwards and Schneider 1997). Thereafter Principles Governing IPCC Work (1998) and Procedures for the Preparation, Review, Acceptance, Adoption and Publication of the IPCC Reports (1999) were formally written for the T A R . Another addition to the T A R was the inclusion of review editors. The review editors were governmental delegates and EPCC Bureau members included as observers and, in certain situations, advisors as scientists drafted the SPM. They were included to ensure that scientists responded to each of the 500-600 comments made in the SPM expert/government review in a fair and legitimate way. These internal observers were included to discourage challenges of bias in the scientific process.  5.4.1 The Policy Relevant Scientific Questions (PRSQ) Moreover the structure of the SYR changed in the T A R . Chairman Watson performed a unique task where he consulted with COP delegations to develop ten policy relevant scientific questions (PRSQ) that were used to frame the Synthesis Report. The ten questions were rolled into nine and were revised and approved by delegations to the COP.  88  The decision to produce a Synthesis Report, that would be based on policy relevant questions provided by governments, was taken at the Thirteenth Session of the IPCC in September 1997 in the Maldives. Questions were solicited by the Subsidiary Body for Scientific and Technological Advice (SB STA) of the U N F C C C and considered by the IPCC and finalized at its Fifteenth Session in 1999 in Costa Rica (Watson et al. 2001, Preface). A core team of 35 lead authors was selected to negotiate the relevance, adequacy and sufficiency of the information necessary to respond to the questions in a way that illuminates the key findings. The addition of the PRSQ's was an attempt to increase the relevance of the IPCC products to the parallel political negotiations in the U N F C C C process in a more explicit way.  89  Table 5.2: A Historical and Causal Typology for the Increased Science-Policy Interface in the IPCC  Phase of the IPCC  Science-Policy  Reason for  Interface  Increased Political  Result  Presence FAR  Includes  Scientists negotiate  To increase  a) Intergovernmental  governments (or  with W M O , UNEP  political saliency  Mechanism  'users') in  and governments for  of the climate  nomination of  a global assessment  issue and  scientists and  with the aim of a  international  review.  Convention.  cooperation.  FAR  Extending the  To ensure  Increased  b) Comprehensive  borders of  comprehensive  relevance to the  Assessments  science to create  assessment  policy sphere  a more policy  especially regarding  through WGIII in  relevant  the costs benefits of  particular.  knowledge  mitigation in WGIII.  continuum.  SAR  Involves  To ensure the  Increased  a) Extensive two  1) external  selection of relevant  scientific  stage review  experts and 2)  information is  credibility of the  experts and  unbiased and  information and  governments  complete.  legitimacy of governments  formal review of W G material.  SAR  Distinguish  Communicating  A separation and  b) Formalized SPM  between the  concise 'take home  demarcation of the  and SYR Process  scientific product  messages' for  boundaries  and the product  approved by  between the  for governments.  international  underlying reports  governments.  andtheSPM's. SYR minimizes W G siloes to provide an  interdisciplinary overview of key findings that are more closely aligned with the COP process and further removed from the underlying report. TAR  Select  Concern in the SAR  Further inclusion  a) Inclusion of  governments  about the biased  of governments in  Review Editors  included in  selection of  order to foster  revision process.  material.  legitimacy of the underlying selection process.  TAR  Integrates  The Chairman's  Inclusion of the  b) The Policy  governments in  attempt to develop a  user community in  Relevant Scientific  framing  closer relationship to  a front-end  Questions in the T A R  scientific  the U N F C C C .  framing of  Synthesis Report  questions.  (SYR)  relevant scientific questions.  5.5 Inescapable Borderlands: Managing the Interface  The shift from the F A R to the SAR and then to the T A R reveals the way that global knowledge is less than effective when it is unidirectionally transferred from the scientific to the policy sphere. Instead the orthodoxies of both the knowledge producers and knowledge users dictate the ways information is interpreted and communicated. The result of the second assessment was to formally distinguish between the science and the non-science in order to deter criticisms of politicization in scientific efforts (see shade differences in Figure 5.1). This did two things, 1) it increased the credibility of the scientific processes due to the sheer volume of participation as well as the validation of the expert review, 2) it created alternate instruments apart from the scientific deliberation and practice where knowledge brokers (LA's, C L A ' s and Co-Chairs) could make selections on what they considered to be the most policy relevant finding through the SPM and SYR processes (see Appendix C for more details on the processes).  91  Figure 5.1: The Structure of the IPCC Reports  SYR S P M rpec Plenary  Synthesis Report Questions SPMWGI Chapter review process TOD Gov't/NG&N Review J  WGI Report  I  SPM W G n I  SPMWCIII  Tech Suinni.  T e c h S uu n mm.  WGII Report  W G III Report  X  SOD Expert Review  FOD Individual Chapters ZOD  * the darker areas with white text indicate policy relevant documents while lighter areas with dark text represent the underlying scientific documents.  The IPCC could not be legitimate while being viewed as elitist, nor could it be credible while viewed as an epistemic network of scientists interested in mitigating greenhouse gas emissions.  The IPCC was  treading a fine-line. Although it was initially intended to be a one-time assessment that would urge politicians to action, the IPCC continued into a second iteration. In this continuation the EPCC not only received support from government-sponsored TSU's (the U K , Japan and the U S took on WGI, II, III respectively in the SAR) but also the IPCC Bureau reflexively re-evaluated (and continued to re-evaluate) the structure of the IPCC. Based on the problems and criticisms encountered in the F A R , the structure and procedures of the IPCC were purified in order for the formal IPCC structure to distinguish the technical and scientific reports from the political documents. Where charges of politicization were mounted, there was now a clear delineation between science and non-science (typical of technical controversies) yet ironically, through this process of demarcation, science-policy interactions increased in the policy relevant processes of the SPM and SYR. The addition of review editors and formal procedures and protocols in the T A R contributed to the assurances of legitimacy in the crafting of the summary document. Yet the scientific activities of the underlying W G reports were distilled and separated from  92  the demands for relevance among nations with competing interpretations of what is and is not relevant. One veteran scientist claims that,  One thing we have learned I would say now is we know where the line lies. I don't think anyone knows it as well as IPCC does. Where does the line between policy and science lie? When I say science I mean expert knowledge I don't only mean just science. I think it's important to know that and not cross that line. For example, avoiding policy prescriptive language is one of the criteria we have to do that (Interview [38], 2001).  In this way the IPCC has derived one great project in managing the science-policy interface (the first element of the Imbued Meaning framework).  The IPCC is a complex adaptive system that has responded to external forces throughout its three iterations. In its effort to retain greater legitimacy, the IPCC has developed the understanding that scientific information from mandated science processes has little practical meaning unless it is valued and validated by the broader socio-political sphere. The IPCC is an institution that is neither a laboratory nor a political organization. Instead it systematically weaves the two together through its complex arrangements and practices.  When counter-experts, developing nations, and analysts looked for and found ideological bias in the underlying interpretations of the FAR, the response of the IPCC was to develop more formal processes in the SAR to demarcate the boundary between science and policy. Moreover, the growing demands to connect the work of the IPCC to policy in the COP processes and particularly the understanding from Article 2 of what constitutes 'dangerous' anthropogenic interference led to the development of the PRSQ's in the T A R to strengthen the connections between the underlying assessments and the policy process. This pendulum swung between the need for scientific credibility and the need for policy relevance. In the IPCC, the governmental sphere is brought into the process in order to internally accept and politically authorize the information produced. The continuation of the IPCC into a fourth assessment report (4AR) that is mandated to be complete by 2007 demonstrates the continued relevance of the IPCC and its importance among governments. It is also worth noting that the attraction of large numbers of veteran scientists and new scientific participants, who are all unpaid, demonstrates a continued relevance and value to the scientists who participate in the international assessment. At the beginning of the T A R , Sir John Houghton, Co-chair of WGI, was quoted as saying,  93  I hope we will continue to get leading scientists. If they get disillusioned with the IPCC process, then the quality of the reports goes down. That will be bad for science and for policy (Masood 1997c). Indeed the IPCC continues to attract notable scientists. The quote above suggests that there is faith in the IPCC process and that the process is capable of producing quality assessments in the underlying reports. However there have been a number of veteran scientists who have either reduced participation in the IPCC or stepped down altogether. This is mainly due to the time commitment required to participate in the process and due to the financial and research opportunity costs (scientists are not paid to participate).  55  The proposed six-year period between the publication of the T A R (2001) and that of the 4AR (2007) was a response to minimize author burnout as participants attempt to balance their professional and everyday lives with the regular schedules, meetings and deadlines of the IPCC. It also allots more time for research projects to conclude and to be included in the next round of assessment. WGI C L A , Sir John Houghton, noted that half of the six years would be spent assessing new information, the other half would be spent collating and writing (personal communication, September 21, 2001).  This type of demarcation plays an important role in 1) maintaining social discipline, 2) balancing authority of science and policy, and 3) ensuring communities are seen to act in appropriate jurisdictions even when, in practice, there is very little a priori distinction between the two at the margins (Miller 2001). These changes were aimed at differentiating science-policy disciplines and demarcating the boundary between science and non-science, while at the same time accommodating the needs of governments for timely and policy relevant information. A more elaborate structure that increased inclusion and enrolment of allies contributed to the stabilization of the IPCC process, its funding and its findings.  5.6 Implications of Analysis Over the course of the first three assessment reports, the IPCC developed processes that have been responsive to external pressures. The IPCC adapted to criticisms to its credibility and legitimacy by instituting changes intended to make it more accountable to external authorities, namely the international policy community. The user community has been included in the review, interrogation and approval of policy relevant scientific representations through: 1) the acceptance of the underlying W G assessment reports where participating governments consensually agree to accept the authenticity, 2) the approval of  For instance, Tom Wigley, a senior scientist at the National Center for Atmospheric Research and long-time participant in the IPCC, decided not to participate in the TAR, due to the time devoted to the IPCC process. He is quoted as saying that his time is better spent" doing science, rather than reviewing it" (Masood 1997, 894). It would be an interesting research project to interview previously involved scientists to determine the range of reasons why they did not continue their participation. 5 5  94  the SPM where governments approve the key scientific findings and their representation on a line-by-line basis and 3) the adoption of the SYR where governments approve of the cross-Working Group synthesis that addresses PRSQ's on a section-by-section basis (the PRSQ's are also approved by governments participating in the U N F C C C COP process). The SPM and SYR processes are boundary processes that contribute to the development of policy relevant products and will therefore be analyzed in detail in the next chapter. For our purposes here the involvement of governments in the process is considered to be an assurance of political "buy-in" to the process (both for the acceptance of existing climate science and for continued funding) while the acceptance, approval and adoption of the text is considered to be an achievement of international "buy in" to the products.  56  These areas represent the first of the five  characteristics involved in the Imbued Meaning framework: a managed interface that preserves the identities of each community.  The formal separation of the scientific assessment reports and the documents of the governments in the IPCC preserve the identity of objective, rational and neutral science - a benefit to both scientific and policy identities. While the IPCC has been generally successful at maintaining distinctions between the work of the scientists in the underlying reports and the participation of the governments in the summary and synthesis documents, it is important to go deeper into the process to analyze the processes and protocols that contribute to that success.  In order to remain policy relevant, the IPCC crafted a space where face-to-face science-policy interactions could take place in order to draft and approve policy relevant scientific information. This occurred in the SPM and SYR. The SPM and SYR processes become the areas where great importance is placed on deriving "policy relevant not policy prescriptive" information. The distinction prefaces or highlights the ways that scientists must follow certain processes and protocols with regard to communicating scientific information. In the following chapter, these processes and protocols will be examined in a more intimate way in order to observefirst-handhow policy relevant scientific information that does not tell governments what to do is produced. Chapter six looks to this arena to understand the ways that science and policy boundaries are managed, transgressed and negotiated internal to the process where the most accessible, popularized and circulated climate findings are found in the SPM and SYR.  This of course does not consider issues of power, coercion, silent dissensus and similar issues in the final governmental consensus. It would be a vast yet interesting undertaking (beyond the scope of this dissertation) to examine individual governmental perspectives toward these processes. 5 6  95  5.7 The Following Chapter The Imbued Meaning framework will be used to analyze the case study of IPCC and the science-policy interface presented through the instruments of the SPM and SYR. Although policies are not made or negotiated in the context of the IPCC, it is widely recognized that the products of the IPCC particularly the SPM and SYR are intended to contribute to domestic policy decisions internationally. The primary research question being investigated and that emerges from the theoretical framework is: In a process where international delegations are directly included in the scientific assessment process through the SPM and SYR, in what ways do the processes, protocols and discourses established in these two mechanisms ensure the credibility and political legitimacy of the information produced? Does increased sciencepolicy interaction in the construction of policy relevant but policy neutral scientific documents make the information more relevant and more salient than if otherwise crafted by scientists alone? Does the lineby-line approval required by delegations increase the collective legitimacy of the SPM information in the international community? These are some of the questions explored in the following chapter.  96  CHAPTER VI: Science-Policy Interactions at the Summary for Policymakers and the Synthesis Report Interfaces In this chapter two facets are examined. First, the three boundary processes that have developed over the past three iterations of the IPCC assessment, that give rise to the policy relevant scientific information in the Summary for Policymakers (SPM) and the Synthesis Report (SYR) will be analyzed. Second, the science-policy interaction in construction of these two boundary objects is analyzed using the four remaining elements of the Imbued Meaning framework. The boundary processes and four elements will be examined simultaneously in the chapter. What will be shown is that the three boundary processes, the review, approval process and the policy relevant scientific questions (PRSQ) provide examples of:  1) A bi-directional exchange and interrogation of discourses and materials across the boundary. 2)  A negotiation of the final representation of the information.  3)  A n inclusion of the 'user' in framing the problem and its technical and methodological requirements.  4)  A n understanding of provisional information that is iteratively examined and reappraised at some point in the future.  What is argued in this chapter is that despite the formal rhetoric in the Intergovernmental Panel on Climate Change (IPCC) that views the managed interface as a unidirectional 'translation' of scientific information into the policy sphere, the formal and informal processes in the SPM and SYR actually create processes that co-produce policy relevant scientific information. In these processes science-policy interaction is managed and new forms of meaning emerge through a two-way process of negotiation and mutual consensus building.  6.1 Participants at the Boundary  The common distinction between a scientific and a policy report facilitates the separation between the scientists who contribute only to the underlying report and those who become knowledge brokers in the document of governments (the SPM and the SYR). These knowledge brokers are the coordinating lead authors.  57  A lead author is responsible for drafting sections of chapters, whereas coordinating lead  authors are responsible for collating and synthesizing the text from all authors in order to draft a consistent, non-redundant and accessible W G chapter. With this overview of the underlying reports, coordinating lead authors (CLA) are also the knowledge brokers who select policy relevant scientific 5 7  There were a few lead authors involved in the final Synthesis Report. How and why they were chosen is unclear.  97  findings that make-up the 10-15 page government documents or SPM's in each of the working groups. Two co-chairs, one from a developed nation and another from a developing nation, oversee each W G and act as intermediaries to the Bureau while also facilitating interactions in the Plenary settings. In this way C L A ' s and co-chairs become active participants in brokering knowledge from the underlying reports to the policy sphere. Some scientists express ambiguous feelings about this position, as the following comment suggests,  Yeah we have to find ways always to make it sound like we're not being prescriptive and that's perfectly fine. But again this is where we have to get into the heads of [governments]. I think what the scientists has been through in this process, they find it very interesting... [Scientists name], I mean I don't want a quote here but [he] was totally disillusioned by the whole thing. Several people here said this is really great... [Scientists name] said this is really great - it's [the SPM] actually a better document. I think what they're saying is that they've actually now had a one week incredible education process about the science-policy event that you almost can't pay for [large laugh] (Interview [42], 2001). The C L A ' s fulfill two roles in the IPCC. They collate and synthesize the work of the chapters in the underlying reports in order to prevent redundancy and ensure consistency. They also select the key findings from the underlying assessment reports for the initial draft of the SPM and SYR. C L A ' s align their efforts with the international policy sphere in determining what is policy relevant and ensure concise, accurate and policy relevant findings are transferred to the policy community.  Providing a summary for governments that is scientifically accurate and approved by all nations is perceived as a way to transcend the boundaries of national politics (Interview [44]). The scientists that work at the interface with policy are 'knowledge brokers' and although they are scientists, they will be referred to as authors (of the SPM and SYR) for the remainder of the thesis. This distinction is made in order to differentiate these knowledge brokers from the participants in the underlying reports. Authors 58  involved in the SPM and SYR are the scientific actors that populate this border territory between science and policy.  The 'user' community refers to international delegations (governments) involved in the W M O or U N E P or both. These delegations are comprised differently depending on the government. For instance, the delegations of some nations are made up of civil servants from the Environment, Finance, Foreign Affairs, or Trade Ministries. A number of members of international delegations participate both in This distinction is for clarity. Participants in the IPCC are all to a certain extent knowledge brokers in that they are doing work that will be used in the IPCC assessment process that transfers policy relevant information to the international policy community even through its underlying reports. The summary products are ways to make the transfer of policy relevant information more concise and relevant summarizing the key working group findings in the SPM. The SYR is another level of summary and synthesis. 5 8  Conference of the Parties (COP) and the IPCC. These delegations represent their nations and ensure that the integrity of the climate information is maintained and presented in a balanced and policy neutral manner. The user community is included in two main facets in the production of the SPM and SYR: 1) through the review process where governments comment on the initial draft of the SPM and SYR, and 2) in the approval process (or the adoption process in the case of the SYR). International non-governmental organizations such as Greenpeace and the Climate Coalition have observer status in the T A R . They are given the opportunity to comment during the review period yet their participation is limited in the approval of the information.. Critical to the success of the SPM and SYR interfaces is that both authors and government participants agree to be involved in the process and also agree with the procedures and protocols that define the process.  6.2 The Role of the Underlying Reports  Participating scientists perceive the underlying reports to be the best and most credible scientific information available. When asked whether authors felt influenced by external forces, most described the underlying assessment process as an "objective" and "decentralized" scientific process and commented that they did not feel any external pressures from governments or the United Nations Framework Convention on Climate Change (UNFCCC) process. Many authors said that the decentralized process of constructing reports and the increased and varied scientific participation actually enhanced the objectivity of underlying information. One author notes that broad participation encourages discursive and cultural understanding in the IPCC assessment process. He comments,  We're here to do the best assessment of the knowledge that's there and we're supposed to do that without reference to our country or our country positions. Obviously we all come with our cultural blinders on and all the rest of it but working in this process over the years I think a lot of those blinders have been lifted for many of us. Not to say that we're perfectly culturally neutral. But I can really understand the arguments of someone who comes from China or some place else. I can see why they've taken the position they've taken or why they think this kind of research is important rather than that kind of research (Interview [32], 2001). The IPCC assessment supports an epistemic community of scientists from varied disciplines who deliberate over materials, interpretations, and areas of conflict and must align their assumptions and practices in a collectively acceptable way. In the formal protocol of the W G reports where agreement is not possible, footnotes are used to highlight research areas that have not achieved closure and/or are controversial.  99  A few scientists noted that this broad participation tended to minimize prescriptive bias in the information. The three rounds of review in the underlying reports also contribute to the perceived impartiality and credibility of the final products among authors interviewed. A WGIII author (with English as a second language) comments,  [What is scientifically interesting and significant] depends on the lead authors interest in viewpoint [and depends] on which view or which paper the lead authors are familiar. Many papers [are] not reflected in the process. Then introduce [the working group report to the] open process then nearly 1000 comments received - some comments [are] on which [missing] literature needs to be assessed. There's so many comments. This kind of review process is very, very good for our work. O f course our viewpoint is very narrow - but expert [review] gives otherwise. 500 reviewer give us comment. Not only one time but two times expert comment and one time government comment. It's very, very good. I can't find such a fair system of integrative process [elsewhere] (Interview [36], 2001). The decentralized process contributes to the feeling of autonomy but as this author recognizes it may contribute to biased or narrow reflections of the literature. However the review process introduces overlooked literature or misinterpretations and strengthens the impartiality of the report. The first external expert review is perceived to highlight missing areas while the second round of expert and governmental review ensures that the information is adequately represented and is consistent. In essence the IPCC has derived its own set of practices and meta-scientific norms that preserves the perceived autonomy of experts that is accompanied by two rounds of extensive expert review. Once the information has passed through the rigorous processes of chapter construction and extensive review it contributes to a sense of objectivity and overall assurances of the neutrality and credibility of the underlying reports among scientists and users of science.  Throughout the IPCC process the underlying reports are perceived among authors to be the realm of objectivity, neutrality and 'fact'. For instance in drafting and revising key policy relevant findings in the WGIII SPM and the SYR and its SPM, comments such as "Are these the facts?" "The proof is in the facts." "It's physics - it's fact!" "It's in the table- it's fact!" were not uncommon (participant observation, L A Meetings, 2001). These comments illustrate the level of trust vested in the underlying working group reports and reveal authors' desire to appropriately convey the facts to the international policy sphere. The processes in which these "facts" are conveyed to the international policy sphere are equally important in order to facilitate approval of the SPM and SYR (discussed later). The critical point is that despite considerable uncertainty, approximation and overall contingency of climate change science, there is general agreement among authors that the underlying reports represent the best available summary of current scientific understanding that needs to be translated to the policy community. The underlying reports are used as a reservoir from which to choose key policy relevant findings.  100  6.3 Boundary Processes in the Summary for Policymakers and Synthesis Report  A n IPCC Bureau member highlights the ways that the IPCC is walking the line between these professional boundaries,  One thing we have learned I would say now is we now know where the line lies. I don't think anyone knows it as well as IPCC does. Where does the line between policy and science lie? When I say science I mean expert knowledge I don't only mean just science. I think it's important to know that and not cross that line. For example, avoiding policy prescriptive language is one of the criteria we have to do that (Interview [38], 2001). The IPCC is grappling with the oft-feared border territory between science and policy in two ways. First, differentiating between the area of science and the area of policy that was formally done in the separation of the underlying scientific reports from the government or policy relevant documents (as noted in Chapter 5). Second, producing policy relevant documents through the SPM and SYR present information in a "concise and accurate" as well as policy relevant and legitimate way. The "line" in the quote above refers to a distinct separation between science and policy in the production of the government documents.  The unidirectional transfer of science into the policy sphere does not work in global environmental issues where credibility and legitimacy is based on trusted and inclusive processes more than on the accreditation of experts (noted in Chapter 5). The IPCC has therefore calibrated and slightly revised the assessment model. In the transfer of information from the underlying reports the IPCC has developed three boundary processes described here as processes that facilitate and manage interaction between the science and policy communities.  59  They are the 1) review, 2) approval and 3) PRSQ processes that  produce the final SPM and SYR products. The SPM and SYR processes were created to ensure that the selection of key scientific findings could be more effectively communicated or translated to the policy sphere (Bolin 1994). These government documents and the associated boundary processes were a response to the criticisms of aligning W G reports directly with the policy community. Within these processes a unidirectional transfer of science into policy does not occur. Instead more nuanced sciencepolicy interactions take place at the interfaces of the SPM and SYR.  6.4 The Review Process The initial exchange between authors and governments is through the review process. Once the individual chapter writing teams craft the first draft, it is circulated via e-mail to experts and participating  This understanding was borrowed from the understanding of boundary objects that translate and exchange materials, discourses and ideas across divergent social worlds (Star and Griesemer 1989). 5 9  101  delegations for review. Reviewers have six weeks to send in both their General Comments, and line-byline detailed comments on the draft.^The T S U collates all of the comments (each consisting of some mix of general and detailed comments, some quite lengthy) in sequential line-by-line order and accompanied by the name of the delegation or individual reviewer. With the collation of the comments into one document, authors are able to see the often multiple and sometimes inconsistent or even conflicting requests for changes on any particular sentence or line. They must differentiate between legitimate and illegitimate comments accepting only those that enhance the overall consistency, balance and neutrality of the text. But all comments must be responded to in writing, thus ensuring that an explicit justification must be provided for either not responding to, or not adopting, every comment made. These responses are written next to the original comments in documents that are hundreds of pages long and provide transparent justification for why the comments are included or excluded by the chapter team (in the case of comments on the underlying reports) or the coordinating lead authors and SYR. writing team (in the case of the comments on the SPM's and SYR). See Table 6.1 for the number of comments.  Table 6.1: Number of Review Comments in Examined Documents Document  Number of Comments  WGIII SPM 1) Expert and Government  550  WGIII SYR 1) Expert and Government 2) Government  Approx. 2352 602  WGIII SYR SPM 1) Expert and Government 2) Government  Approx. 1424 490  In some cases the process of responding to the comments is simple and clear. For instance, some comments indicate where minor editorial or grammatical changes or oversights are required, such as a missing footnote or the placement of a clause or the replacement of a word to emphasize the relative strength of a message. In other cases the issues are less clear-cut. For example, language in the form of legal and political discourse becomes a significant focus in the review process where proposed grammatical and editorial changes are made in order to communicate scientific messages in a way that is perceived to be impartial, balanced and non-prescriptive. Language such as "will", "should" and "can" is identified as too prescriptive thereby infringing on the territory of governments to determine policy goals and outcomes. There is a resistance among many nations against this binding discourse. Yet other governments who support precaution and urgent mitigation may see this language as necessary. Beyond the neutralization of the language there are other informal exchanges and interactions that take place.  102  6.4.1 The Bi-Directional Exchange of Materials and Discourses  The review process fits with the bi-directional interrogation and exchange of materials, the second element of the Imbued Meaning framework. There are a number of things that occur through the review process: 1) governments use particular frameworks to interpret and modify information in the SPM, and 2) these governmental comments sensitize authors to the values and interests of different governments. In the government comments and their reactions to particular language what becomes tacitly revealed are the different ways that the scientific message can be emphasized or de-emphasized based on the framework of the government reviewing it. A simple example of this is provided in the statement, "Development paths that focus on the social, economic and environmental elements of sustainable development may result in lower greenhouse gas emissions" (participant observation, WGIII SPM, March 1; emphasis added). One nation suggested that the word "will" replace "may" with the ending "in all countries" (WGIII Comment 86 p.22). Another nation recommended inserting "or may not" after "may" to indicate the trade-offs associated between economic, environmental and social justice concerns (WGIII Comment 87 p.22). In many cases what at first glance appears to be an editorial comment can at second glance be identified as a recommended change that emphasizes or de-emphasizes a particular message. The former comment emphasizes the role that development paths play in reducing greenhouse gas reductions while the latter emphasizes an ambiguity in the message that is less prescriptive to policymakers. Authors, in this case, concluded that "may" inherently captured the ambiguity and "will" was rejected as too policy prescriptive. This example reveals how contradictory comments may sometimes play off each other, resulting in the retention of the original text.  Another pattern in the review of the SPM were particular words that caused controversy among governments. Authors referred to these as "nuclear words". These words include such terms as "sinks", "targets", "precaution" (insinuates mitigation), "inform" or "require" that can be linked to controversial concepts, or that suggest some form of binding obligation, in the COP negotiations. These words and at times, concepts were, removed or replaced by authors drafting the SPM in order to minimize contestation. Knowledge of these politically charged terms enabled veteran scientists to make judgments about whether the potential for destruction in the intergovernmental approval process outweighed the relevance and importance of the proposed text (destruction is referred to as the breakdown of the consensus negotiation process). Many lead authors are veteran scientists who have been through more than one IPCC process. These authors have been imprinted with the types of language and "nuclear words" considered prescriptive by governments. This sensitization is revealed through lead author comments observed in the process such as "Is it better to put it in now and to get hung up in the approval or to remove it now?" (participation observation; L A Meeting, June 20, 2001).  103  A Nuclear Example Highlighted in the review are contradictory understandings of what is and is not policy relevant and policy prescriptive among governments. As noted above, authors are required to address and integrate, these often-contradictory comments if they are not demonstrably wrong. Governmental review editors observe this process to ensure that authors legitimately deal with all comments. For instance, internal document analysis reveals considerable concern among governments with nuclear power being identified as an alternative energy in the initial draft of the SPM WGIII (IPCC doc 3).  60  The inclusion of nuclear as  an alternative source of energy had the effect of inciting a number of nations to contest this sentence stating that it was an inappropriate energy alternative under the objectives of development, equity and sustainability (DES).  61  Authors' attempts to facilitate governmental concerns are constrained by what has  been assessed in and is consistent with the underlying reports. Authors thereby had to defend the text due to the fact that nuclear technologies are considered to be a viable energy alternative in the underlying report. The three to four authors examining this chapter agreed that "nuclear is an important carbon-free technology operational in many countries and should not be removed" (WGIII doc. 3, p. 2, line 3). However in order to integrate the concerns of opposing nations on this issue, authors constructed an additional sentence. This sentence was approved as, "Economic, environmental safety and proliferation concerns may constrain the use of some of these technologies" (Metz et al. 2001). What is interesting in this additional sentence is that in authors' efforts to integrate governmental concerns about nuclear power, they anticipated that it would be important to also mention other alternative technologies. In other words, isolating the social and environmental concerns of nuclear without highlighting the cost concerns of other alternative technologies could be identified as an imbalanced representation of the underlying reports by governments (and non-governmental observers such as F O R A T O M ) who were in support of nuclear as an energy alternative.  62  Authors must anticipate these reactions in order to minimize controversy and delay  in the line-by-line approval of the SPM.  What is revealed is that beyond a governmental review and 'translation' of scientific materials, the review process in practice is an indirect exchange of information both from authors to governments and from Using document analysis, the areas of response from authors are identifiable in the review comments. Areas that require broader consultation and discussion among lead authors and co-chairs and are highlighted by the phrase " L A ' s to discuss" and highlight areas in the representation where there is ambiguity and interpretive flexibility among authors. Both F O R A T O M (European Atomic Forum) and Greenpeace were external reviewers that contributed opposing comments on the nuclear line. There was an interesting juggling act that authors' were involved in when integrating government comments that were not demonstrably wrong. How to integrate comments became tricky because the other governments that did not comment because they agreed with the way the sentence was worded. It did not necessarily mean that the sentence was not critical to them. For this reason, there were times when authors' integration of say, four government comments on a particular line introduced greater dissent in the Plenary because the remaining 120 noncommenting governments had approved of the original message. 6 0  6 1  6 2  104  governments to authors. The review comments highlight the divergent positions, biases and interpretive frames used by different nations and scientists in determining what is a consistent, balanced and policy neutral representation. They expose the ways that particular nations and delegations view particular words, language and content and thereby highlight potential areas of controversy for the approval process. The review and interrogation of the messages occur within a controlled process ensuring that charges of imbalance and inconsistency are dealt with in the confines of the boundary process.  The nuclear power example illustrates three things are simultaneously occurring in the review process: 1) governments can have contradictory comments and interpretive frames, 2) in responding to governmental concerns authors are guided by what exists in the underlying reports, however 3) veteran authors realize that they must adapt governmental comments and concerns in ways that will minimize conflict in the approval process. Each community is constrained in what they can and cannot say based on the consistency with the underlying reports. With consistent involvement in the process, scientists are able to tease out these political codes to try to deflect the level of controversy in the line-by-line approval of the SPM (Interview [44] [42], 2001). Therefore reflexivity is required, on the part of authors, to anticipate what is and is not a consistent and balanced statement.  6.4.2 Reflexive Feedbacks  The critical scrutiny provided by governments in the review process also helps to sensitize authors to bias in their own frameworks. The language used, the representation of information and concepts, and the underlying prescriptions identified in these representations come under critical scrutiny. One WGIII author identifies the ideological bias he had in his first SPM participation,  Essentially you cannot present a biased opinion - you always have to think back to be as neutral as possible and this is the kind of learning that takes some time. I remember my first participation I had not these views so I was just trying to explain what it was like for developing countries and so on. And then I learned that it is necessary to maintain a balanced position (Interview [35], 2001). This author identifies the type of reflexive learning that occurs with continued participation within the process. What authors can and cannot say must be based on consistency with underlying reports and balance of message. Authors must negotiate among the varied and contradictory comments to determine the most appropriate way to address each comment whether contradictory or not. This is a significant undertaking that influences the revised SPM draft that is then circulated in the Session of the Panel for line-by-line approval. Based on previous experience in the process and on the potential controversies 105  highlighted in the review comments, authors collect information that enables them to anticipate and forecast difficult areas for the approval process. For instance, when discussing Figure 7 in the SPM of the SYR which illustrates the technological possibilities for early versus delayed action and the respective mitigation costs, a veteran author notes that "economists will say that cleaner technologies will only occur with GDP growth; the US will hate it because they are top emitters and China will like it because they have lower per capita emissions" (participation observation, June 22, 2001). These anticipated trade-offs are reflexive feedbacks of the process that illustrate tacit learning among authors in their efforts to draft a scientifically accurate and yet approvable summary document.  With the information learned from previous experience and in the review comments, authors have become quite clever in presenting information (as identified in the nuclear power example). What can and cannot be said is the project of the scientific interactions in disciplines and sub-disciplines that contribute materials to the IPCC. As one author notes,  [The authors] have become quite sophisticated within their discipline in being able to decide and agree on the message that they want to convey so that what you find in WGI while there might big disagreements within a discipline or sub-discipline is that some of the LA's can effectively be ambassadors to that discipline and they would've had a meeting and they would've agreed - what it is that was acceptable to be said and what it was that was acceptable not to be said or not acceptable to be said (Interview [32], 2001). This author identifies a learned sophistication among authors in their ability not only to represent information in a way considered acceptable but also in reaching agreement on what is and is not acceptable to say, what may cause controversy, and what should be omitted altogether. The interrogation by governments influences the ways information is interpreted and represented, what to include and exclude, what will be acceptable and unacceptable and therefore contributes to a bidirectional exchange of materials and discourses between governments and authors. This becomes particularly pertinent at the science-policy interface where determining what can and cannot be said is not only determined by authors but involves the interpretations of governments as well.  6.5 The Approval Process The approval process is where the revised SPM (from the review) is formally introduced into the intergovernmental process called the Session of the Panel. The Session of the Panel "refers to a series of meetings at the plenary level of the governmental representatives to the IPCC" (IPCC 1998). The Panel (or Plenary) is the critical science-policy interface where the revised draft is approved by governments  106  and where the diverse and often-divergent interests and values of nations become transparent.  63  In this  process the common epistemic goal among the majority of authors and governments is to achieve consensus approval of each line of the SPM.  In the Sixth Session of the IPCC Working Group III held in Accra, Ghana 124 governments were represented by delegations comprised of one to seven  Table 6.2: Number of Participants in the WGIII SPM Approval  officials (see Table 6.2). 40% of attending governmental  Delegations (124)  No. Participating  U K and China. International organizations were allowed to  1-2 Officials  109 nations  contribute to the review but were relegated to observer role  3-4 Officials  6 nations  during the approval process. The approval process took  5-6 Officials  4 nations  place in a UN-style forum in a large room, with horseshoe-  7 or more  5 nations  officials (49 in total) were from Japan, Canada, the US, the  shaped seating arrangements for governments supplied with microphones, nation titles, and earphones that translate  Delegates  Organizations  21  Lead Authors  25  interactions into the five non-English U N official languages (Russian, French, Spanish, Chinese, Arabic). The Chairman sat at a panel at the front of the room with the Secretary to  his left and the relevant lead authors to his right (the authors switch according to their area of expertise). A 15-foot screen expanded across one side of the front panel and participating nations examined text for approval in the case of the SPM or adoption in the case of the SYR. The proceedings took place in English and generally opened with a keynote address from an official of the hosting nation. In this forum the question became, what is a credible and legitimate document that can be agreed to by international governments? What is consistent, balanced and policy neutral for the diverse interests and values of the international community? This question became critical in this UN-style forum where each nation is equipped with veto authority over the final message. The role of authors became transformed in this process where in order to achieve approval and ownership by governments, consensus had to be achieved 64  on the way scientific information was represented.'  Pre-emptive efforts to prevent stalling in the approval process are made by Bureau members prior to or during the approval process. For instance, the Chairman mentioned that in drafting the SYR SPM, China and the US were the only governments to comment on ways to better address issues of "dangerous" in Question (and Chapter) One. They were taken aside in order to communicate the status of the research and to come to a conciliatory agreement before the Plenary. The approval process is very similar in the SYR SPM. However in this SPM approval process, by cutting and pasting previously approved information from the WG SPM's, authors attempted to bypass the considerable controversy of the approval process. (This is discussed in more detail below). 6 3  6 4  107  6.5.1 The Negotiation of Underlying Frameworks in the Approval Process  The line-by-line approval process contributes to a face-to-face interaction in the Plenary that enables differing governmental interpretations to be made transparent and for final representations of scientific information to be negotiated in the larger Plenary. The ways information is represented becomes critical to governments who are able to represent information in multiple ways while remaining within the confines of what is technically correct.  In the Accra plenary, this interpretive flexibility was illustrated in the negotiation of the ways the costs of stabilization of greenhouse gas concentrations in the atmosphere at 550ppmv should be represented. In the underlying WGIII report, this cost had been expressed using two different but synonymous measures. The costs of stabilization were represented both as 0.06%, a proportion of the global economy, and as 2% of the gross domestic product (GDP) of governments. There was deliberation among authors about the best way to communicate the cost of stabilization in the SYPv (participant observation, L A Meeting, June 22, 2001). The small proportion of the global economy was agreed to. Yet, in the approval process, this particular representation came under critical governmental scrutiny. The US argued that representing the cost as 0.06% of the global economy did not provide an adequate picture of the costs and made them appear inconsequential. They argued instead that the absolute cost quantities would provide a better reflection of the costs of stabilization (i.e. trillions of US dollars). The negotiation began when other nations argued that the metric in trillions made numbers look very large and over-emphasized the costs of mitigation.  While a small number of governments supported the suggestion, nations that had been content with the original representation argued against the proposed changes of the US, thereby stalling consensus approval on this particular line. In attempting to address the stalemate, authors' were prepared to make the alternate suggestion to represent the costs as a proportion of GDP instead as both representations were consistent with the underlying science. After considerable negotiation and pleading for cooperation from the Chairman, an average of 2% GDP loss was approved, revealing the simultaneously small and yet considerable fraction of GDP required to stabilize greenhouse gas emissions at 550ppmv (participant observation, June 22, 2001). This example illustrates the ways that the representation of information can change the emphasis of the underlying message while remaining technically accurate and consistent. In the Plenary forum, these types of negotiations also had the indirect effect of highlighting governmental (and author) preferences over ways to represent climate information and which areas to emphasize or deemphasize.  108  What occurred in the SPM was a complete review and revision by governments with different interpretations of what was considered to be a consistent, balanced and neutral message. These are not objective criteria that give rise to unambiguous wording choices. Since the underlying reports are expansive enough to justify multiple interpretations, in the approval process governments negotiate messages that comply with or are based on their particular frameworks, for instance regarding the types of information considered valid. At this interface negotiation is not necessarily demonstrative of a search for an accurate word.' Rather words are connected to national interests, values and conceptual frameworks. The credibility and legitimacy of a claim cannot be achieved by reference to technical knowledge alone but must have processes of communication, exchange and negotiation incorporated in order to discern what is acceptable or in the case of the SPM, approvable, by the policy community.  6.5.2 Negotiating Valid Information  What became apparent in the approval process is that different nations sought different types of information that coordinated with their underlying value frameworks. For instance, many developed nations sought quantitative facts and technical information and viewed qualitative methods and information as policy prescriptive. When referring to the controversy and negotiation in the approval process a WGIII author notes, "Its not the language, it's whether information can be pinned down to quantitative facts" (Interview [30]), illustrating the importance powerful governments attach to quantitative information.  The type of information sought by a number of O E C D nations was made clear in the WGIII SPM review comments. One nation commented that, "The SPM strikes me as being too weak on numbers and factual information, although such information exists in the underlying report" (New Zealand, WGIIII SPM, Comment 17). Another nation was concerned that,  The SPM gives an unduly optimistic picture of the potential for greenhouse gas mitigation in the short term, and gives a misleading impression that greenhouse gas limitations/reductions can be achieved largely at little or no cost... This is an important issue which in our view is not supported by the evidence in the full Report, and a more nuanced picture is needed that takes full account of the uncertainties (Australia, WGIII SPM, Comment 44). Implicit in this latter comment are concerns about the underlying bias in the information that does not sufficiently focus on numbers, costs nor adequately reflect underlying uncertainties in the scientific information. Due to what are perceived to be a lack of technical facts, the US states, "it [the WGIII]  109  appears to highlight or advocate specific perspectives that may dilute its overall credibility as an unbiased reflection of the state of the knowledge.. ."(US, WGIII SPM, Comment 30). The overall sense that there was a lack of policy neutrality in the SPM was based on certain conceptual understandings about what constitutes legitimate neutral science. Differences in these understanding gave rise to a number of controversies in the approval process (discussed in more detail below).  These review comments foreshadow the particular concerns that emerge in the approval process. For instance, an IPCC Special Report on Emission Scenarios (SRES) undertaken in support of the T A R , projected six "plausible" future worlds, comprised of varying combinations of economic growth, technological progress, population growth and governance structures. These constructed futures were used to determine the influence of socio-economic drivers on greenhouse gas emissions and greenhouse gas concentrations in the atmosphere that could feed into the WGI general circulation models (GCM) (Nakicenovic and Swart 2000).  65  The SRES scenarios are analytical devices that are less about deriving  technical facts and more about making explicit the various assumptions and drivers involved in different development paths. They highlight the ways that future choices of governments and citizens influence emissions patterns (refer to Figure 6.1) and they can contribute to broader contextual thinking about the role of development, equity and sustainability (DES) in determining appropriate responses to climate change. However due to the subject matter (socio-economic, technological, governance change and decision-making) and their proximity to policy, these scenarios cause considerable scepticism among governments seeking predictive and quantitative 'facts'.  In the T A R , governments requested a transparent expression of uncertainties with each scientific claim made. Uncertainty lexicons were produced in WGI and WGII however attachment of probabilities and uncertainties posed a challenge for the types of information assessed and produced in WGIII.  66  For  instance, authors of the SRES Report (2000) that produced varied projections of future greenhouse gas emissions argued strongly that no probabilities could be attached to the different scenarios since each one  In the TAR, the IPCC contributed to the production of new climate information through the SRES scenarios. These particular scenarios made the IPCC an explicit creator, rather than only an assessor, of knowledge. Such a role for the IPCC has an even longer history in climate science. Although the GCM's used in WGI are constructed in different international research programs it is reasonable to suggest that they are constructed mainly for use in the IPCC. In this way the IPCC has also been a driver of WGI models along with other climate information. Many authors commented that the advantage of participating in the IPCC is to identify gaps for future research (Interview [27] [30] [33], 2001). Based on this analysis policy relevant research is identified by IPCC participants and due to its relevance is more easily supported by national governments. There was also considerable concern over the ability to attach uncertainties to WGII impacts information (participant observation, LA Meeting, June 22, 2001). However lexicons for expressing uncertainties in the scientific information were developed for claims made in both WGI and WGII. Inconsistency between the WGI and WGII uncertainty lexicons was identified and highlighted as problematic by Bureau members in the drafting of the SYR (participant observation, LA Meeting, June 19, 2001). 6 5  6 6  110  Figure 6.1: Climate change - an integrated framework  Taken from the IPCC SYR SPM 2001.  represented a different set of internally consistent social and individual choices and decisions, which could not be compared in terms of relative probability (personal communication, Robinson 2004). This supported the claims made by some government's who criticized WGIII for being too normative and policy prescriptive, not scientific. Many of the controversies that were sent to contact groups were based on contradictory understandings about what types of information, quantitative or qualitative, were considered to be policy neutral and valid in the context of the WGIII SPM approval process.  67  A Bureau  member confirmed this observation when he stated that, "All the normative statements here if [governments] don't like them they're considered [policy] prescriptive... [and] die a miserable death" (Interview [42], 2001).  While developed nations tended to ask for more quantitative information, many developing nations questioned the legitimacy of technical information and required that qualitative concepts be included. For instance, China's concern (highlighted in the general review comments) does not focus on a lack of  A Bureau member addressed this problem in another way stating, "Simply WGI and WG II are risk assessment and WGIII is risk management - that puts it right on the edge of the policy framework therefore there's more potential for being criticized [as policy prescriptive]" (Interview [42]). 6 7  Ill  technical and factual information in the WGIII SPM but instead identifies a technical bias in the SPM draft. The Chinese government comments,  In this SPM, the main impression is that there is large technical potential both in the near-term and long-term to reduce annual global greenhouse gas emissions at 2000 level and achieve stabilization of atmospheric concentrations in the range of 450-550 ppmv. In fact, due to great difficulties in removing social and market barriers related to technology development and transfer, it is impossible to realize the technical potential. In addition it should be noted that since 1990 little progress has been made to reduce greenhouse gas emissions in major developed nations, which should take the lead in mitigating climate change as required by the Convention (China, WGIII SPM, Comment 4; emphasis added).  This comment highlights China's concerns regarding the optimistic focus on technical potential for reducing greenhouse gas emissions. Rather than focussing only on technical aspects, China highlights the need to focus on social and economic barriers that prevent such potential. What is at issue here is the view that a technocratic focus has inadequately addressed the more social and cultural variables such as access to markets, social and cultural appropriateness and equity considerations that contribute to the effective transfer of technical possibilities.  68  Issues of DES were considered to be relevant and necessary  material by many developing nations. This was largely due to equity considerations that ensured a historical examination of development patterns, systemic inequities, and feasible future development possibilities (as noted above). In both the review and approval processes it became apparent that developing nations question the scientific and technical focus due to the lack of social and historical content. For instance, India charged that stabilization scenarios focus on economic optimization that does not adequately address issues of lifestyle, consumption patterns and per capita emissions (participant observation; Plenary, March 6, 2001).  In viewing these contradictory requirements among nations, authors develop an awareness of the sensitivities of certain issues through the approval process. This was expressed in interviews with authors where many were able to rattle off the divergent concerns of governments stating that the U S was interested in costs, the U K in precaution and developing nations in equity (Interview [30] [32] [36] [42], 2001). Reflexivity about these issues was observed as authors drafted the SYR and its SPM, one author questioned, "Could this get through the plenary? You must think about this as you craft sentences". Another author responded "we were hammered on society, behaviour, values, social learning - they were Much of this technocratic focus links back to the policy instruments negotiated in the COP process such as the Joint Implementation (JI) and Clean Development Mechanism (CDM), largely based on technical and technological transfers to developing nations otherwise known as "leapfrogging'. For instance, the estimated costs for biological mitigation in the TAR WGIII SPM range from $0. l-$20 USD per ton of carbon in several tropical countries and from $20-$100 USD in non-tropical countries (Metz et al. 2001, 26). Projected costs are projected to be less expensive in the developing world, which has lead to policy instruments such as the JI where developed nations can collect carbon credits for helping to reduce carbon intensity in the developing world. 6 8  112  omitted as too sensitive" (participant observation; L A Meeting, March 3, 2001). The omission of such information in the approval process as too policy prescriptive or "too sensitive' is indicative of how the review and approval are not only about the language and better communication of scientific information but encourages the negotiation of value-frameworks among governments (while also providing guidance to authors about policy relevant research in the future). The schism between the types of information sought by different governments and the validity of concepts was central to many of the conflicts in the contact groups (participant observation, WGIII SPM and SYR SPM, 2001).  Finding different ways of expressing scientific information that considers the contradictory concerns of particular governments became the project of all participants in the approval process (not just the authors as in the review). It was largely the effort of other governments and even more significantly the Chairman that contributed to mediating many conflicts in the larger Plenary. However when the process fell apart, at times due to destructionist behaviour, mechanisms for conflict resolution were initiated. 69  Where consensus became cumbersome, the Chairman would order particular problems over messaging to be resolved through smaller satellite "contact" groups. Governments, authors, Bureau members had a common goal for all lines to be approved in the SPM document and all sections to be approved in the SYR document. If this could not be done in the larger plenary than concerned governments and relevant authors crafted new text in contact groups at breaks (and due to time pressures at the end of the week in multiple groups that ran parallel to the Plenary).  6.5.3 Contact Groups  In these informal sessions, a government official would chair the meeting at the front of the room. The line or paragraph under scrutiny is projected through a projector onto a screen and Microsoft "track changes" is used to edit text according to the deliberations and negotiations among the delegates and scientists in the room. In these contact groups, individual actors (both scientists and delegates) are involved in dialectic interaction over appropriate wording and messages in a less formal setting. What  Many authors commented on the destructionist behaviour of the Saudi Arabian delegation (Interview [27] [30] [44], 2001). My observations confirmed that from the beginning of the approval process to the end, the Saudis argued for changes to be made to the majority of lines under examination. Many authors saw this as a mis-use of veto authority in the process. However a number of Saudi comments made in the review, suggest that their concern about the messaging in the SPM was largely due to the assumption made in the IPCC that stabilizing greenhouse gases at lower concentrations (450 and 550ppm) would minimize catastrophic effects. They viewed the majority of research based on how to stabilize greenhouse gases at these lower concentrations as skewed toward an emphasis on reducing concentrations. Without much if any research examining effects at 750 to lOOOppm, the underlying assumption of reduced GHG was labelled as policy prescriptive and therefore undermined many of the projections that were the basis of IPCC findings. An international focus on reducing greenhouse gas emissions, places a significant proportion of the Saudi economy at risk.  6 9  113  was noticeable in these sessions was that authors were able to describe the science in more detail and delegations were more able to express their concerns. In one situation an author was able to describe the assumptions in the economic models in more detail so that governments had a common understanding of the information at hand and were based on this additional information were able to come to agreement on the message. In another meeting a delegate announced, "If I agree to this, my head will be cut o f f while authors and delegations helped to minimize the concern (participant observation, Contact Group, Sept. 25, 2001). These informal exchanges that take place between governments and authors reveal the informal nature of the contact groups. In this setting, each participant, whether author or delegate, is participating in crafting text that can be approved by all disputing parties. What was observed in this setting were micro-negotiations that encouraged informal exchanges between participants with less of a focus on the boundary and more of a focus on crafting textual representations that are approvable to each participant in the contact group. Contact groups continue until text is written that is acceptable to all parties in the group, and is then approved in the Plenary.  70  Contact groups were in general not viewed as desirable due to the time and effort required to reach agreement. In fact the Chairman used contact groups as a threat in the approval process; "Please accept this wording or it will be sent into a contact group." Despite this, these informal sessions occurred fairly often. Every other page would have an issue that was causing delays and problems in the larger forum that would be sent to the contact groups. Contact groups cumulated toward the end of the week meeting in the early mornings, at breaks and at times, during the Plenary.  Example of a Contact Group  The issue of equity raised contradictory interpretations in paragraph 4 in the WGIII SPM (participant observation Panel, March 4). The issue of equity was discussed in the plenary for close to forty minutes before it went to a contact group. The US in particular viewed equity as infringing on the role and jurisdiction of government. The US challenged that the issue was neither scientific nor technical and was  The work by CLA's, LA's and Co-Chairs that goes into drafting the policy relevant documents is considerable and yet at each stage of the review and approval a complete rewrite is required. An anecdote provided by a WGIII author reflects the reality of the effort to achieve consensus in the SPM. 7 0  We had an amusing incident in a contact meeting we've been working in. The chair, who is a delegate of course, had got something together and then the comments came in last night and now they had to be changed. And we met with the chair this morning and he said, "How can they tear down this beautiful language we've put together" and we all [the authors] started laughing. He asked why we were laughing and we said "Well you sound like a LA" (Interview [34], 2001). The Chair of the contact group experienced only a minor version of the frustration that authors experience regularly in this process. 114  therefore policy prescriptive in the context of the approval process. However a number of nations viewed the issue of equity as paramount. Nations such as India, China, and the U K highlighted equity as a legitimate scientific issue (discussed in the underlying WGIII report) that provides analytical understanding to the climate change problem both from a historical perspective and for a future decisionmaking perspective. Referring to Chapter Four on Equity in the WGIII report, the U K delegate argued that, "if the IPCC raises questions of equity in a significant way it would be a good starting point... valuable" (participant observation, Plenary, March 4, 2001). Paragraph 4 was sent to a contact group that spanned the course of two days and numerous breaks in the Plenary.  In the contact group the US argued that there could be no global consensus on "fairness". They argued that what is "fair' to the developing world such as having industrialized nations diverge from the 2008 baseline in the first commitment period (2008-2012) is not considered "fair" to those nations assuming those additional costs. They argued that the additional costs imposed on developed nations to diverge from the 1990 baseline earlier than developing nations must be considered when determining what is and is not fair and made the suggestion to replace the concept of equity with marginal costs of abatement thereby placing an emphasis on costs.  71  The US governments concept of equity was informed by  economic factors different from the historical responsibility and inequitable development patterns that concerned many developing nations (including Kenya, China and India). The US governments' effort to equate equity with economic costs can be viewed as an attempt to transform a qualitative and "unscientific" concept into a quantitative and what was perceived by them to be a policy neutral concept. However what this controversy illustrates is that the US suggestion to use quantitative cost estimates in order remain policy neutral exacerbated rather than deflated conflict over issues of policy neutrality. The focus on quantitative estimates represented a particular framework focussed on cost-optimization that did not consider different possible frameworks focused on lifestyle, consumption, historical and current per capita energy use for instance.  One observer highlights the problem with a focus on qualitative issues such as DES in the approval process, Some of the statements that have come in and disappeared are more what many delegates would call "belief statements" which I think we all believe in... well not all of us some of the delegations don't, but if you actually go and look at the literature you find that they're just belief statements they're not actually statements about what has been observed in the world.... I mean for example the development, equity and sustainability  The subtext of the lines of argument over equity run parallel to the controversial issues of differentiated responsibility in the COP process. The tenor of many controversies in the approval process including the one on equity mimics the ongoing political negotiations. 7 1  115  issue. The statement says that if we have more equity, more development and more sustainability then emission mitigation might be better. Now that might be true - we'd all like to think its true - you can't actually show it (Interview [32] 2001). Issues such as development, equity and sustainability (DES) involve normative questions of fairness and equity that, in the approval process, are considered by a number of powerful nations to be less robust or less scientific. "Belief statements" are not viewed as sufficient ways of directing policy trajectories and are therefore considered to be policy prescriptive by a number of governments seeking predictive and quantitative statements. The concepts involved in DES were viewed as normative and prescriptive because "it insinuates some kind of governmental or social change" (SYR SPM, Saudi Arabia, Comment 56). However inequitable development patterns, and related historical responsibility are issues central to the involvement of developing nations and according to that framework require further scientific analysis.  72  A Bureau member comments how the inclusion of issues such as equity become critical in  order to engage the international policy community in cooperative efforts. He states,  The other thing with global environmental issues is with climate change or with ozone depletion or persistent pollutants or whatever, internationally you do have to have a concerted coordinated international action and you do have to recognize equity distributional issues, you have to recognize who's already screwed the environment up, you have to recognize stages of development. A l l of these things are critical otherwise it's not gonna go forward-just not gonna go forward. Therefore our assessments have to bring all of those dimensions in (Interview [44], 2001). The emphasis here is on coordinating international value frameworks and the respective assumptions underlying them relating to issues of responsibility, relationships to contemporary institutions, lifestyles, consumption and development paths in order to move forward. These normative issues become just as critical as the selection of technical information in determining an adequate and approvable policy relevant scientific message.  In the review and approval processes, both the values in the science and the diverse values of governments are simultaneously negotiated. In the review governments identify and interrogated underlying bias or areas that are not policy neutral in the science. Where they are considered to be unbalanced or policy prescriptive, governments must negotiate new representation considered to be more appropriate while remaining within the bounds of being consistent with the science. Governments have  A significant part of the controversy over equity was the limited assessment of peer-reviewed research that had been done in the WGIII underlying report. In the contact group the US identified the main source of equity research as Toth (1999) and challenged it as a single source reference not sufficient to be included in the SPM. Whereas, developing nations and a number of authors argued that it had been through a considerable expert and government review and was therefore sufficient to be included (participant observation, Contact Group, March 4, 2001). 7 2  116  differing interpretations of what is unbalanced and policy prescriptive. For instance attempts to express the impacts of climate change on developing nations was seen by the US to be imbalanced due to the fact that it focussed only on the negative impacts of climate change and not on the potentially positive impacts. Similarly, discussions of mitigation were perceived by some nations to be policy prescriptive and to assume, and therefore prescribe, a collective policy response to climate change (US, WGIII SPM, Comment 36). It was therefore requested that the concept of adaptation accompany discussions of policy response alongside mitigation highlighting both as policy options. One WGIII author agrees, "Mitigation and adaptation must be a balanced portfolio. For instance developing nations may see the urgency to adapt more than the need to mitigate" (participant observation, L A meeting, June 19, 2001). The consistency and balance of information is important in order to offset governmental needs.  6.5.4 The Approval Process: A Negotiation of Words or Concepts?  Although there is considerable science-policy interaction and negotiation, and, at times, compromise in the approval process, there was still a resistance among some veteran authors to view it as such. For instance, a veteran Bureau member cautions against perceiving the science-policy interactions in the approval process as political negotiations. He asserts that,  The negotiation is only how to clothe the uncertainties and gaps in knowledge and things of that kind. It looks like negotiations. It's really not political negotiations in the true sense of the term. Any group of experts, when two people write a scientific paper together, they would argue over words. This one is an order of magnitude more argument than [the co-authorship of] a paper, given the fact that all of this has policy implications] and governments are fully aware. But the fact that [delegates] make an effort to stay away from straight policy, we have to say hats off to them for doing that, just as experts struggle not to cross that line into policy. I think it's a good process. It's a fascinating process (Interview [38], 2001). The emphasis here is on translation and the negotiation of words. This Bureau member stresses the separation of governments who stay away from "straight policy" and experts or authors who struggle "not to cross that line into policy". This interpretation of the process is based on the formal IPCC rhetoric that perceives the SPM as a translation process where climate change science can be more effectively communicated to the policy sphere. A number of other authors had a different view of the role of governments in the process. A WGIII author highlights the difference between the underlying reports and the SPM. He states that, "They [governments] don't try to tamper with the underlying documents but they certainly try to tamper with the message that appears" (Interview [34], 2001). This author recognizes the negotiation in the approval process as that of the governments attempting to realize their own national interests or positions in the message. A WGI author emphasizes this point by focussing on the changes  117  that the SPM document goes through; ".. .And the whole [SPM], even words and sentences, is scrutinized and approved. And [the government officials] debate over it. And so the whole thing actually gets rewritten in the final meeting" (Interview [44], 2001). The suggestion here is that in the translation process governments tamper with scientific message and rewrite the entire document based on governmental negotiations and national interests.  However the negotiation is not entirely that of the governments negotiating interests and positions. A WGIII author suggests that the ways to qualify the message becomes a project of C L A ' s as they consider the highly charged setting of the Plenary. His depiction of the SPM is that of authors' framing messages in order to be as neutral on normative issues as possible while political delegations continue to pull them back in. When asked whether the process is based on consensus or negotiation, this author responds,  It's [the SPM] negotiation - what level are we going to caveat this - what are we going to say about it? And then toward the SPM we try to make sure that none of the normative stuff is there but of course some of the delegates want to pull some of it back in again. But that's fine - the SPM is the policymaker's document and we are just the drafters of it. On the other hand I think we've all learned not to suffer pride of ownership but rather to recognize that this is a very political document and we are trying to bridge between the substantive analysis and the political desires of the countries (Interview [34], 2001). What this author identifies are the ways normative issues are brought into the process. In this account the idea that the SPM is a bridge between science and policy that connects both substantive analysis and political desires reflects the idea of the SPM as a hybrid. Scientists write the initial draft and frame the substantive analysis while in the review and approval processes governments pull in the diverse frameworks of participating governments.  Authors' references to the negotiations in the approval process either focus on the negotiation of words or the negotiation among governments in securing a legitimate and normative message. These descriptions of the approval process support the understanding that the SPM is a political document. Yet if the SPM were entirely a political document it would not have the legitimacy and "buy-in" of international governments. What is not readily highlighted in these characterizations of the approval process is that the authors and governments are both constrained by what can and cannot be said in the approval process. Author revisions must be made within the bounds of being consistent, balanced, neutral and relevant in order to retain the legitimacy of the underlying reports. Governmental revisions must be within the bounds of being scientifically correct and balanced. The SPM is neither a political document nor a scientific document; it is some hybrid of both.  118  Moreover in the efforts to maintain a unified position on what can and cannot be said in the SPM, underlying imbalance and controversy can be masked. The negotiation of what information is selected and interpreted for a governmental audience can be controversial in itself. For instance, one WGIII author refers to the scientific representation in the SPM, "I know of a number of authors that did not believe what they were presenting but did so anyway" (Interview [27], 2001). A few authors highlighted the considerable differences in economic research between what one-author referred to as "AngloSaxons" and "the rest of the world and commented on the dominance'of neo-liberal economic ideology in Chapter 2 and throughout the WGIII report (Interview [33] [27], 2001). Free enterprise, consumption and individualism are the underlying assumptions in the majority of the economic literature and thus the economic models that originate from O E C D nations. A WGII author notes that the process "isn't balanced between north and south" (Interview [30], 2001). He continues,  In fact it is Anglo-Saxon dominated. French, German, Russians are [also] disadvantaged because virtually all literature is assessed in English. North- South inequities exist simply because of where a lot of research is done and resources and manpower from particular regions of the world. These are things I don't think any process could correct (Interview [30], 2001). Another author comments on the western as opposed to northern dominance in the IPCC process, "The US can veto because it is well-organized and well-published. It's not their fault, it's our fault" (Interview [33], 2001). What is emphasized here is not only differences in methodology among WGIII authors but ideological, interpretive and framework differences in the underlying assessment. The IPCC protocol to use only peer-reviewed literature is an effective way to ensure that credible information is assessed.  73  However the majority of scientific journals and peer-reviewed literature is based in O E C D  nations which, with regard to the dominance of economics in' WGIII, skews the assessed information in favour of the neo-liberal ideology dominant in those areas of the world.  74  Another author asserts,  "Consensus does not accept this ideology" highlighting that this "ideological economics is an imperialism that spreads into other nations" (Interview [27], 2001)  7 5  This makes it difficult for developing nations or  nations with different political and socio-economic platforms to actively enter into the process.  76  What is  emphasized here is that the negotiation in the SPM is not only over words but is the negotiation of  It should be mentioned that there are formal procedures for the introduction of non-reviewed literature (IPCC 1998), though such literature remains a small proportion of that included in the IPCC reports. The research literature is heavily dominated by the US largely with the respective underlying commitments to neo-liberal marketization and globalization (Interview [27] [32] [33]). Yet these two sceptical authors continue to participate in the IPCC process. One says he continues to participate not due to an environmental ideology but due instead to a concern for the future of mankind (Interview [33]). It is important to highlight these issues of implicit power and authority in order to develop a more accurate understanding of the interactions throughout the process. However detailed analysis of these power differentials is beyond the scope of this project. 7 3  7 4  7 5  7 6  119  appropriate literature and underlying concepts that frame the options available for policymakers. These concepts are based on the underlying assumptions, commitments and value-frameworks embedded in the science (see Appendix D).  Although governments negotiate the SPM representation this is not the same type of negotiation one would expect in the U N F C C C forum. Instead governments are in part negotiating the underlying values, assumptions and messages that are embedded in the science and that give rise to varied governmental interpretations. Governments negotiate the policy relevant message in order to find ways to politically neutralize (or neuter) the tacit codes in the scientific message as much as possible in a way that 'fits' with different governmental frameworks. The underlying scientific prescriptions, that would otherwise likely be accepted by some nations, and rejected by others in a unidirectional model of science, are negotiated in the approval process in order to achieve consensus on what is considered to be a politically legitimate and appropriate scientific message. This negotiation has trade-offs. Instead of strong prescriptive statements about how to mitigate climate impacts for instance, the WGIII SPM provides what some authors refer to as "obsequious", "more complex" and "more complicated" statements (Interview [30] [32] [42], 2001). For instance, the projected impacts and costs of climate change on developing nations cannot be discussed without simultaneously highlighting the benefits of a 2-3 C average temperature increase on the economies of developed nations. A Bureau member comments on the problem of integrating governments in the approval of the policy relevant scientific message. He states,  Some people are questioning whether there should be an IPCC because they see the IPCC as either being too biased, as being too politicized, as not really understanding the rules of different regions. As an example when we say that the poorer countries need not do something then the others immediately say hey you're against us. [An attitude of] he who is not with me is against me is developing (Interview [38], 2001). 77  This is one of many examples of the way that scientific messages undergo negotiation in the intergovernmental sphere in order for tacit codes, preferences and prescriptions to be neutralized. In other words, the underlying biases, codes and prescriptions implicit in the scientific messages are reviewed, interrogated and negotiated by governments in the review and approval processes.  This same Bureau member continued to discuss the perceptual bias in the scientific and expert causes problems for the IPCC in the media. He comments, 7 7  community that  So we still have perceptual problems. The perceptual bias comes because there are still vast number of scientists and experts who for whatever reason question IPCC findings. If they publish their work in the literature then we could look at that but if they just talk to the media then we don't know how to deal with that (Interview [38]).  120  6.6 The Policy Relevant Scientific Questions The purpose of the SYR is to "synthesize and integrate materials contained within the Assessment Reports and Special Reports and are written in a non-technical style suitable for policymakers and address a broad range of policy relevant but policy neutral questions. They are composed of two sections as follows: (a) a Summary for Policymakers and (b) a longer report" (IPCC 1999). The policy relevant scientific questions are a new addition to the T A R SYR process. The PRSQ's have been referred to as "the Chairman's baby" (Interview [28], 2001). He contacted a number of nations asking what issues they cared about and asking them to draft questions for scientists' to address. The Chairman describes this initial process,  Yes I went to SUBSTA, I went to the COP process. Through S U B S T A which is the Subsidiary Body on Scientific and Technological Advice, I said what do you care about what are your policy issues now? They could never at S U B S T A tell me. So we finessed it brilliantly. I got the American government, Small Island States, the European Union all to think about it and they sent it [the questions] to us through the S U B S T A process. So this is not S U B S T A as an entity... But the governments in SUBSTA, the E U said this is what we care about. The American government said this is what we care about... So we then knocked out the politics of it and we came up with the 9 questions and we took it back to them and directly they said, "yeah that captures the issues we care about". So that sort of played it backwards and forwards from the policy to the science process (Interview [n/a], 2001). 78  There was a considerable back and forth until participating nations to the Subsidiary Body on Scientific and Technical Advice (SBSTA) approved the questions. This effort focussed on greater attempts to further connect science and policy. The significance of the PRSQ's is that they facilitate an explicit inclusion of governments in the framing of what is policy relevant. This is different from the SPM's that have authors select and communicate the key scientific findings considered to be policy relevant. However what is scientifically relevant may not be policy relevant. Rather than having authors anticipate governmental needs, governments make their needs explicit through a series of questions in the PRSQ's. The Chairman describes the benefit of making this link,  Most National Academies [and] many scientists have great ideas for assessments. We can get together but if no one's waiting for our report it's just gonna collect dust. So you almost want to bring [policymakers] into the process, get them involved in the scoping, what issues do they care about. So what happened in this last document [the SYR], the one we were just talking about here, we asked the policymaker's what were  The [n/a] refers to the fact that the interviewee has been identified. Inclusion of the interview identification number would expose this person's identity throughout the rest of the research and is therefore withheld. It should be noted that the Chairman has been identified in areas of his initiative in the development of the process. This develops understanding about intention of the processes. His identity remains concealed elsewhere. 7 8  121  their nine questions they would take home. So they're waiting for the document (Interview [ri/a], 2001).  The PRSQ's included policy expertise and sensibilities at the front end of the process in order to guide the selections and interpretations of authors from the underlying reports and Special Reports. In this regard, the SYR is an arena where the combined expertise of scientists and policymakers both become pertinent to framing and producing policy relevant scientific information. The idea is that including governments at the front-end heightens interest and sensitivity in the reports and thereby ensures an audience for the final product.  Selections for the SYR were made by a team of authors who were agreed to by the Bureau, noting the need to aim for a range of views, expertise and geographical representation (IPCC 1999). Each question was considered based on which working groups would be required for addressing the question and authors were assigned from those working groups. Authors communicated via e-mail during the construction of the report with infrequent face-to-face meetings including the meeting I observed in Annapolis, M D .  Each of the 35 members of the core writing team were veteran lead authors and had therefore been sensitized to the political discourse, sensitivities and considerable negotiations of SPM approval processes. This was important due to the fact that the underlying SYR would go through a similar process referred to as the adoption process. The SYR information undergoes simultaneous expert and government review and is thereby scrutinized on a section-by-section basis in the Session of the Panel. This process brings the SYR information under more scrutiny than the underlying W G reports and less scrutiny than the line-by-line approval process in the SPM. The section-by-section adoption of the underlying SYR required similar political savvy in representing information that addressed the policy relevant questions. All three W G reports and the five Special Reports were used to address the policy relevant scientific questions in an attempt to align the IPCC more closely to the COP processes. For instance, Question One addresses the role of science and the IPCC findings in determining what constitutes 'dangerous' anthropogenic interference. 'Dangerous' is a highly normative term with considerable political implications that refer directly to Article 2 of the U N F C C C . It is therefore critical that authors be familiar with the process. The Synthesis Report is also accompanied by an SPM that represents a condensed synthesis of the policy relevant information. Chairman Dr. Robert Watson took over 100-page report and extracted what he believed were the key messages into a 26-page summary (14,000 words) (IPCC communique, 2000). He put this document out as a call for the writing team to let him know which messages are essential and which can be dropped in order to write a 5-7 page SPM.  122  6.6.1 Including the User in Framing What is Relevant  The PRSQ's represent an innovative method for increasing the policy relevance of the scientific information in the international policy community. Whereas the W G SPM's encouraged the transfer of key scientific findings and the negotiation of their representation, the PRSQ's are an effort to integrate the user community in framing questions that are considered to be policy relevant information. Rather than having authors anticipate the needs of policymakers (which is an indirect feedback of the process), the SYR encourages governments to explore what is policy relevant. This transforms the material from that which is scientifically interesting to that which is policy relevant. This is important in the crafting of policy relevant scientific information. One author notes, "Collaboration is seen as a weakness [in the IPCC]. The SYR is a response to that..." (Interview [27]). The questions were intended to guide authors' selection of information to address these relevant questions.  In drafting responses to the PRSQ's, authors were encouraged by Bureau members and co-chairs of all three WG's to use quantified information and figures as much as possible. The Chairman argued, "The key to success for the Synthesis Report and the SPM will be punchy take home messages and thoughtful tables and figures" (SYR communique, April 22, 2001). The underlying emphasis was based on the understanding that more text equals greater governmental input into the core writing process and therefore contributes to more work (participant observation, L A Meeting, Sept. 17, 2001). Efforts were made to reduce the workload on participating authors as much as possible. A Bureau member notes that, "In reality it's all going to be out of WGI and II. We're just cutting and pasting as you can tell, but a bit more than that, a bit of synthesis but I wouldn't spend time on it" (Interview [44], 2001). In practice, the level of interdisciplinary synthesis that took place across the WG's was minimal with the majority of the writing process being based on cutting and pasting previously approved information. WGI and WGII information was preferred due to the fact that probabilities were attached to the scientific claims. WGIII information experienced considerable contestation in the processes (for reasons noted above) due to the fact that many nations considered its materials normative and policy prescriptive. Authors used procedural strategy to bypass political controversy. Part of the strategy was based on the Simple Rule used in the IPCC. The Simple Rule states that previously approved information cannot be reopened for 79  negotiation. A Bureau member makes this strategic point; "It's almost better [to take text] from the SPM because it's been agreed to by policymakers" (participant observation, L A Meeting, June 19, 2001). In the SYR and SYR SPM there were a number of times, where a delegate took issue with particular text  The title 'Simple Rule' was used informally among participants in the process to describe the protocol that prevented previously approved informationfrombeing re-opened to review. 7 9  123  only to find that it was directly taken from an approved W G SPM and was therefore not available for revision.  What was observed is that the complexity of both the political and interdisciplinary process was bypassed in favour of procedural ease. Without a formal deliberative process involving COP governments in the negotiation of pertinent questions and informational requirements, the questions became more open to criticism. As a result both science and policy participants claimed that the questions were poorly constructed and largely ineffectual (personal communication Robinson 2004). Although the PRSQ's generated an innovative approach that attempted to include the user community in determining what is most pertinent and relevant in the policy world, authors' previous involvement in the Plenary approval processes contributed to a response focused on ease rather than relevance.  6.6.2 The Example of "Dangerous Anthropogenic Interference"  At the Annapolis meeting of the Core Writing team a Bureau member identified "dangerous" and "uncertainties" as considerable issues in the SYR. Question one directly refers to the language of the United Nations Framework Convention on Climate Change (TJNFCCC). It asks, "what can scientific, technical and socio-economic analyses contribute to the determination of what constitutes dangerous anthropogenic interference with the climate system as referred to in Article 2 of the Framework Convention on Climate Change"?  Implicit in the request for scientific information to contribute to determinations of 'dangerous' is a • request for a scientific characterization of a highly normative concept. In the construction of the SYR and its SPM in Annapolis, M D , authors discussed a number of ways to frame and address this question. These ranged from one author suggesting that the word 'dangerous' be defined for the SYR glossary to another author expressing concern over whether to address such a normative and prescriptive concept within the process (participant observation, L A Meeting, June 18). It was soon agreed that in defining the term dangerous, the context-specific nature would be lost and would cause significant controversy in the SYR adoption process.  Author's concluded that "dangerous" means different things to different nations and different people. For instance authors discussed that the Alliance of Small Islands States (AOSIS), the regions projected to have the greatest impacts and the most to lose, view contemporary anthropogenic interference in the climate as dangerous. Whereas governments with increased adaptive capacity perceive the term and the threshold of dangerous much differently (participant observation; L A Meeting, June 18, 2001). Adaptive 124  capacity refers to "the ability of a system to adjust to climate change (including climate variability and extremes) to moderate potential damages, to take advantage of opportunities, or to cope with the consequences" (Watson et al. 2001, glossary). It was concluded that the term "dangerous" depends on value judgments associated with the projected impacts of climate changes and the varied socio-political processes such as adaptive capacity to deal with them (participant observation; L A Meeting, June 18, 2001). Another author noted that dangerous does not only correspond to regional differences or a specific threshold in space and time but also involves the dynamics of the rate and magnitude of climate change (participant observation; L A Meeting, June 18, 2001). This understanding shifted the discussion from regional contexts and adaptive capacities to an increasingly global emphasis on mitigative capacity. Mitigative capacity is defined as "the social, political, and economic structures and conditions that are required for effective mitigation'''' (Watson et al. 2001, glossary). This concept moved the term dangerous from being a construction of a local decision-making environment to a concept that focussed on the overarching social, political and economic aspects that either deter or contribute to globally coordinated possibilities for reducing anthropocentric interference of climate change. It was concluded that higher mitigative capacity implicitly reduces the danger of a given level of anthropogenic interference. Both adaptive and mitigative capacities were highlighted as areas where scientific, technical and socioeconomic analyses could contribute to determining what constitutes "dangerous".  In the expert and governmental review a number of governments highlighted a wariness of the term 'mitigative capacity' and then again in the Panel Session. An external expert comment provides an example of the type of critical reaction that arose in the review and approval process to the inclusion of mitigative capacity in Question One. A US expert reviewer comments,  Has the IPCC assessment process established that the climate change problem REQUIRES C O L L E C T I V E global determination and actions? While this seems plausible for maximized effectiveness, the link to where this point is elaborated upon in the T A R (or other IPCC findings) would be important to cite. Otherwise it could be misperceived that the Synthesis Report is advocating collective policymaking. For example, a "collective requirement" would seem to somewhat contradict the "acting individually" statement... (Expert, N O A A Aeronomy Lab, USA). The insinuation is that mitigative capacity focuses on collective determinations and actions to abate increased anthropogenic interference on the rate and magnitude of change. This expert anticipates the way an emphasis on collective decision-making may be interpreted as advocating a specific response in the international policy sphere. There is increasing recognition that emissions reduction will not necessarily lead to reductions in global greenhouse gas (GHG) concentrations and that, due to lag effects, impacts are inevitable and will require strategies for adaptation. However, mitigation of greenhouse  125  gases is generally not recognized as an adaptive response strategy in the way that it is pre-emptive. The German delegation commented,  .. .assessing the impact of the magnitude and rate of climate change on inter alia ecosystems, human settlements, agriculture, economic development... is independent of the assessment of mitigation policies and of the mitigation capacity (see Questions 5 to 7). These cannot determine what is a dangerous climate change (German Gov. Comment 31). Mitigation strategies are perceived by a number of nations to imply global decision-making. This becomes a critical area of discussion in the adoption of the underlying SYR and the approval of the SPM. As the US delegate succinctly states: "Adaptation is about risk assessment which can be better understood through research whereas mitigation is seen as risk management and is thereby the jurisdiction and territory of the political. [The US] can't support this statement [mitigative capacity] because it is too prescriptive" (participant observation, Plenary, Sept. 28, 2001). The US was referring to the fact that mitigative capacity addresses the social, political, and economic structures and conditions required for effective mitigation thereby stepping into what was viewed by the U S and other nations as the soft and prescriptive arena of governments. In this definition there are normative assumptions about the role of governments in responding to climate change. Climate change is researched as a global phenomenon and there is an implicit assumption that the response of governments will be collective and global as well (i;e. the Kyoto Protocol). A few governments identified this normative imperative in the term mitigative capacity and resisted its inclusion. According to these nations', how to appropriately react to climate change is the territory of legislatures, policymakers, and policy analysts, not IPCC scientists.  Moreover, the government of Denmark referred to mitigative capacity as an entirely new construct that was being used inappropriately with regard to determinations of dangerous; "What level of abstraction? [Mitigative capacity] is not appropriate in this forum where we deduce risk. I think you should ask what governments want not what you want" (participant observation, Plenary, Sept. 28, 2001). Relating the crafted text back to the question being addressed, the WGIII Co-chair argued that Article 2 of the U N F C C C "has a timeframe aspect that effects the socio-economic [and it is therefore] legitimate to include mitigation" (participant observation, Plenary, Sept. 28, 2001). Question One emphasizes that the term "dangerous" does not correlate to a particular point in time nor does it have a homogenous meaning, leading to a dynamic rather than static understanding of the term. Given that greenhouse gas concentrations cumulate over time (and have considerable lag effects) there are unknown variables that increase the possibilities for surpassing thresholds (leading to elements of climate surprise and irreversibility).  126  A number of delegations argued in support of the inclusion of mitigative capacity by arguing for instance, that, "[its] inclusion is based on technical and scientific considerations not on socio-economic judgments" (Australia delegate 2001). The foundation of the argument was explained as follows: "The pathway to stabilization at every level is important to address adaptation. The extent of [dangerous] impacts is not only affected by the endpoint but [also] by the pathway" (participant observation, Australia, Sept. 28, 2001). This statement suggests that two-pronged adaptive and mitigative capacity response to Question One does indeed stay within the bounds of the scientific if one considers the ways that future emissions and development pathways are influenced by mitigation. The U K delegation supported this representation suggesting text that read: "since both rate of increase and absolute level are important .. ."mitigative capacity should be included  Oparticipation observation, Plenary, Sept. 28, 2001). This  relates to a statement also made in Question Seven of the SYR that states, "Both the pathway to stabilization and the stabilization level itself are key determinants of mitigation costs" (Watson et al. 2001, 28). A n interesting procedural note can be made about this negotiation in the Plenary. Both the Australian and U K representatives were involved in the SYR as review editors; one observed the construction of Question One. That both agree with the explanation provided by the authors may or may not have been influenced by this role. Perhaps their exposure as review editors to the scientific deliberations that went into the inclusion of mitigative capacity increased their understanding of the explanation (and/or increased their trust in the authors themselves).  The Chairman also argued for the inclusion of mitigative capacity stating,  The unanimous view of all 20 lead authors supports mitigation. We thought it was a different use of mitigative, not a whole new construct... this is intended to be a scientific process to present facts from construct and this has been done through peer review: two expert reviews and government reviews. Only two governments made comments, the US and Denmark, and neither commented over the past six months... Fire protection does not relate to the risk of a fire in this room - that's fire insurance. Protection is mitigative (participant observation, Plenary, Sept. 28, 2001).  The Chairman uses the credibility attached to the review process and the agreement of all L A ' s as a reasonable justification for including the concept. He relates the concept of dangerous to fire protection. The suggestion is that one does not have to only rely on fire insurance to prevent the risk of fire but can take alternative steps to protect oneself from it (i.e. clean the chimney). The role of mitigative capacity in contributing to protection against dangerous human interference is illustrated in the synthesized Figure 6.1: Climate change - an integrated framework (refer back to page 134). The square box in the figure provides new preliminary insights (addressed in the TAR) into the ways that socio-economic development pathways can contribute to slowing the rate and magnitude of climate change. This take home message from Question One states, 127  The basis for determining what constitutes "dangerous anthropogenic interference" will vary among regions, depending both on the local nature and consequences of climate change impacts, and also on the adaptive capacity available to cope with climate'change. It also depends upon mitigative capacity, since the magnitude and the rate of change are both important (IPCC T A R SYR 2001, p.38 [emphasis highlights change]). This statement suggests there is a need to focus both on adaptive and mitigative capacity. The former is required to reduce vulnerability and increase resilience of biophysical, human and socio-economic systems to increasing greenhouse gas concentrations and the respective radiative forcing. The latter is required to simultaneously minimize future concentrations in order to slow or reduce the rate and magnitude of climate change and its impacts. The inclusion of mitigative capacity closes the loop in the diagram by including the socio-economic drivers that contribute to emissions and influence the rest of the cycle. The inclusion of mitigative capacity relates strongly to the understanding that socio-economic analyses can contribute to the determination of "dangerous anthropogenic interference" and is considered by some nations to impinge on sovereign governmental decision-making relating to development.  This example illustrates the resistance by a number of nations to a prescriptive bias in the underlying science that requires negotiation and collective approval rather than balkanized efforts to address climate change. A Bureau member comments on the controversy of Question One. He states, "A lot of people are seeing that we are placing climate change in the border context of sustainable development. If they were good research managers out there, they'd say, 'my God we haven't spent any money on this" (Interview [44], 2001). Over the iterations of the IPCC assessment reports increased attentions have been paid to socio-economic analyses in WGIII. As the detection and attribution have become increasingly accepted, the overall emphasis of the IPCC has shifted from getting the science 'right' to the focus on the socio-economic analyses about ways human systems and drivers can have less impact on the ecological systems while ensuring equitable and sustainable development pathways.  For the strong co-production of policy relevant scientific information, the user community must be engaged in the problem definition and in determinations of informational requirements at the front end so that scientists are able to collate and assess scientific information that is directly and transparently related to what is policy relevant. Implicit in these two concepts lies the recognition that both technical and normative information respectively are required to derive meaningful policy relevant scientific information. The PRSQ's in the T A R SYR make an attempt in the right direction. Their derivation was based on an effort to make a stronger connection between the U N F C C C policy process and the IPCC scientific assessment. However two problems were identified with the PRSQ boundary process: 1) the 128  questions arose in an ad hoc fashion where a few nations drafted questions considered policy relevant, therefore a rich dialectic among governments about the most pertinent and relevant informational requirements did not occur; and 2) the PRSQ's were addressed by a scientific assessment that was already complete, therefore what governments considered to be policy relevant had little to no bearing on the selection of materials and literature made in the underlying reports.  80  Without explicit informational  requirements and policy relevant questions framing the front-end of the assessment, the ability for crossW G synthesis is hampered by scientific allegiances and hierarchies. This was observed in the SYR where an attempt was made to address policy relevant scientific questions in an interdisciplinary way by cutting and pasting previously approved climate (WGI), technical (WGII) and social sciences (WGIII) information from the W G SPM's to address the questions. The IPCC working group reports were aligned along scientific lines and categories not policy relevant ones, making interdisciplinary integration and policy relevant synthesis cumbersome and time-consuming in the SYR. In the Imbued Meaning framework, in order to construct strongly co-produced policy relevant scientific information, the user community is included to debate, discuss and determine what policy relevant at the beginning of the assessment process and must be reappraised at a determined point in the future.  What is Policy Relevant Changes Through Time  Iterative reappraisal is important for strong co-production. What is policy relevant at one point in time may not be relevant at another. This is also the nature of environmental sciences. The dynamic interactions between the social, conceptual and material worlds change through time The IPCC 81  assessments are iterative. As one author notes, "The IPCC will continue for as long as governments find it useful" (Interview [27], 2001). The iterative nature of the IPCC assessments ensures that every five years new climate change literature is assessed and reviewed. A number of scholars note that agreement on definitions and concepts is easier in negotiations where political actors know that the information agreed to will be reappraised in the future (Eckley 2000; Fisher and Ury 1998). To what extent this has been critical in the IPCC has not been sufficiently examined here to make a claim however it is suspected that the iterative nature of the approval processes contributes to consensus agreement among governments and authors. Although procedurally more efficient, the Simple Rule not to open previously approved information for renegotiation, became problematic for a number of nations (specifically in the  If governments had been involved in deliberating over what is policy relevant before the initiation of the WGIII underlying report there may have been a greater inclusion of literature on the influence of equity issues in climate change mitigation. This is best illustrated through Kuhn's (1962) scholarship that examined "scientific revolutions" where conceptual and social factors contribute to paradigmic shifts. 8 0  8 1  129  controversy over "a few" in the approval of the SYR SPM). As one delegate noted, "just because it's been approved does not mean it cannot be improved upon" (participation observation, Belgium Sept. 18, 2001). The reappraisal of information contributes to strong co-production by ensuring that what is policy relevant (and the information that can contribute to it) is continually reappraised and that relevance is conceptualized as temporally and spatially dynamic rather than static. This is similar to the concept of adaptive management. In adaptive management, human, biophysical and socio-economic systems and dimensions are appraised as they influence and interpenetrate one another through time.  The SPM and SYR have been subject to iterative reappraisal and will continue to be as long as the IPCC is still in operation. The SPM and SYR are reappraised over each assessment report where new information is assessed and contributes to cognitive expansion in thinking about the problem (see Figure 6.1). Iterative assessment also highlights the provisionality and dynamic nature of global scientific information. For instance, the range of average surface temperature has changed from a projected 1.4 4.5 C increase in the F A R to the 1.4 - 5.8 C increase in the T A R (van der Sluijs et al. 1998). Iterative assessment enabled information to be reappraised. Achieving agreement is made easier under the auspices that the information underlying the agreement will be reappraised in the future with the addition of new information (Eckley 2000). In the SPM and SYR boundary objects, that which is approved as policy relevant in one assessment may change in a future assessment. A n unanticipated finding with regard to reappraisal in the boundary processes was the way the Simple Rule in the IPCC prevented it. Not bringing previously approved information under reappraisal and re-negotiation fosters an inability to improve upon or add to policy relevant information.  6.7 Authors' Varied Perceptions of the SPM and SYR  The effort to draft policy relevant scientific information requires the participation and interaction of both scientific and policy communities. The boundary processes provide ways for materials, discourses and values to be exchanged and communicated via formal negotiation. In other words, the SPM and PRSQ processes facilitate interaction and communication between science and policy communities. It has been established that the SPM and SYR innovations arose out of the need to generate political legitimacy and relevance in the climate information produced in the IPCC. However political legitimacy is not the only goal of the IPCC; rather attempts to achieve political legitimacy while being based on defensible science is the overarching goal. Therefore it becomes important to understand how authors' perceive this interaction and negotiation and the ways it affects the credibility of the policy relevant scientific information. How do authors perceive this interaction? Is there a rhetorical understanding of what is occurring in the SPM or are there diverse perspectives among the interviewed authors? What are the  130  patterns, if any, among interviewed authors' with regard to perceptions toward the processes underlying the SPM and its role?  6.7.1 Individual Perspectives Many of the authors interviewed viewed these boundary processes as contributing to international consensus building. Many authors viewed the SPM and SYR as 'government documents' initially drafted by authors but negotiated and approved via considerable political negotiation. A WGI author expresses this by stating, If you look at the IPCC reports and the SPM's that it says something like, "a text as negotiated by the government based on the draft initially prepared by..." So it's very clearly a text of the governments and it is just a way of acknowledging the help with the initial draft (Interview [45], 2001). Another author states, "we've all learned not to suffer pride of ownership" referring to the fact that authors are just the initial drafters of these reports and that political desires must be bridged with the substantive analysis (Interview [34, 2001]). The majority of authors agreed that the review and approval processes provided sufficient checks and balances over what could and could not be said about the science in way that remained credible to authors.  In general the SPM was viewed as a scientifically credible document that is also imbued with diverse and divergent political meanings. A WGIII author comments that, "there's so many political discussions but it's O K - it's important to affect such a scientific knowledge to the policymaker's process" (Interview [36], 2001). Another WGIII author states, The reason you see 78 representatives of the countries here [is that] they are trying to influence the way the document is written. For me it is good evidence that these [SPM and SYR] documents will be used in the future to set directions for policy (Interview [35], 2001). A WGIII author commented on the difference between the underlying reports and the SPM stating, "[These negotiations were] purely political considerations based on phraseology and how it might be interpreted in and influence the negotiating process in the future" (Interview [30]). Each of these authors viewed the negotiations with governments as political in nature and necessary in order to "affect" "set" and "influence" policy.  Suggesting that the authors agreed with the process may be too strong, certainly there were identified trade-offs associated with such negotiations. A few authors viewed these negotiations as a loss of 131  scientific control in the approval process. One WGI author stated, "I really don't like the process and from my standpoint I would prefer to see the scientists with a bit more control" (Interview [45], 2001). This author viewed the negotiations in the approval process as "political interference" (Interview [45], 2001). Another WGIII author suggested that the focus on "language more than translation" (Interview [33], 2001) meant that divergent government views toward appropriate language had to be accommodated, and that the diverse and divergent meanings over phraseology resulted in "convoluted" messaging in the SPM (Interview [30], 2001). The underlying perspectives among these authors were that the negotiations over the interpretation and representation of scientific findings were political negotiations based on how governments could influence the parallel policy negotiation process in the future. This was seen by some as political interference for the quality of the message produced.  A few authors also questioned the nature of consensus in the SPM and SYR. Having to incorporate divergent political meanings, discourses and perspectives contributed to a significant amount of controversy in the process, which was viewed critically among authors. A WGIII author viewed the accommodation of divergent perspectives as leading to compromise, coercion or forced consensus among governments. He argues, "You can say that Mr. A and Mr. B represent opposing views, yet they must come to a level [in the approval process] even if the government is forced. It is badly constructed. [This] is impossible (Interview [33], 2001). The concern for this author was that the scientific information being used was skewed in favour of "an Anglo Saxon ideology" (Interview [33], 2001). He was uneasy about both the quality of the consensus attained and also the quality of the information and message 82  represented. Another WGIII author was critical of the consensus building process stating that,  The meetings could be made to be more consensual and more productive and to make better use of time... What's the alternative? Delegates know exactly what they're doing here. Going late into the night seems to be expected. This is the way things are (Interview [27], 2001). The disagreement and controversy associated with the approval process meant long days were spent debating the underlying meanings of words. These meetings generally began at 8 am and ended, on average, around 10 pm at night. Part of the consensus building and negotiating process was based on stamina (Interview [27] [30], 2001).  83  Despite these criticisms and trade-offs the majority of authors  Lahsen (2000) investigated the nature of negotiation in the IPCC and found differences between active participation and latent participation among actors. She argues that consensus of the final IPCC products does not consider the limited participation among actors from developing nations namely Brazilian scientists thereby bringing the notion of scientific consensus into question. At one point in the SYR, the translators had stayed their maximum time and were required to leave for the evening. The US government suggested that the meeting continue until agreement had been achieved while the Russian government argued that if the meetings continued in Russian they would consider staying on but otherwise they would not compromise (participant observation, Plenary, Sept. 28, 2001). 8 2  8 3  132  supported the inclusion of governments via the boundary processes and were in support of the final product. For instance, one WGI author succinctly offers that, "The SPM is a legitimate document. It is modified and decontextualized but is scientifically correct" (Interview [39], 2001). A number of authors suggested that the review and negotiation (in the approval process) actually improves and/or benefits the SPM (Interview [27] [45] [46], 2001). A WGI author notes,  The SPM always turns out to be better than the draft. The process does work! The SPM has government ownership. They aren't allowed to write things unsupportable by text. It is concise. There is no negotiation. The SPM focuses on tone and emphasis (Interview [40], 2001). In this authors' view, the approval process is not based on political negotiation but is based on more , effective ways to craft the language in the final scientific message. This author perceives the approval as a unidirectional translation process where words not concepts or meanings are being changed.  Interestingly, I interviewed two IPCC participants (each had contributed to the underlying report WGIII; one was a lead author (LA) scientist on a chapter and one a reviewer) with the same unidirectional understanding of science-policy but with opposite views about what was occurring in the review and approval processes. From this understanding, the science-policy interaction and negotiation in the SPM was viewed as particularly problematic. The WGIII scientist illustrates this in his comment,  Science is fact. Policymakers must trust scientific experts and take up the facts rather than promoting positions. IPCC should end at fact! In the SPM facts are being altered. Scientists aren't defending facts adequately. It's not just one man's work, it's all authors' contribution and delegations are skewing the facts (Interview [43], 2001). This scientist is concerned that the inclusion of the governments in the SPM enables governments to politicize the facts. In this unidirectional view of science-policy where science is transferred directly to policy, the negotiation of scientific representations among governments is perceived to be a politicization of science. This observing scientist does not have faith in the processes that underlie this negotiation but instead sees governmental delegations negotiating or skewing the facts and scientists facilitating this rather than ardently defending the original scientific representations. As a L A in the WGIII underlying report, he experienced a feeling of betrayal that his work was not being defended adequately in the approval process. The observing reviewer has a similar unidirectional view and suggests that what comes out of the approval process is not a scientific product. His concern is that the overall representation of the SPM is misleading in the public sphere and argues,  133  Much of what goes on in the SPM is dictated by the IPCC - it is the intergovernmental panel and you can't forget it is governments making the ultimate decisions and those governments are guided by much more than just the technical information. So what gets put out to the public - which is portrayed as a scientific fact - is science strained through the filter of policy and again that is the problem (Interview [28], 2001). Although this reviewer views the process from a unidirectional science-policy model, he has a more nuanced view of what is occurring in the SPM. He recognizes the SPM as a significant part of the intergovernmental process and due to this, the role that governments play in contributing to the representation of the information. His assertion that "science [is] strained through the filter of policy" is portrayed negatively as a politicization of the science and his view that governments make the ultimate decisions about the SPM message feeds this negative view. Neither of these scientists was participating in the WGIII SPM approval process but instead they were observing from the sidelines. Is it coincidental that these observers had much more definitive reactions toward what was occurring in the approval process? It is interesting that both had definitive views about the unidirectional transfer of science to the policy sphere and that information thereby viewed the interactions and negotiations in the approval process as a politicization of science. A governmental official expressed an alternate view of. He claims that the complex processes of negotiation and approval make it "difficult to turn your back on IPCC because of the SPM. What's a better solution than making science accountable" (Interview [29], 2001). This comment illustrates the implicit commitment, whether symbolic or practical, of governments involved in the SPM process. He views the underlying processes in the SPM as a way to ensure that science is made accountable to the political context in which it is operating in order to "tell policymaker's what they want to know but in their context" (Interview [29], 2001).  Except for the one WGIII scientist who was observing the process (Interview [43], 2001), every author interviewed believed that the review and approval processes were the best available ways to perform the task of transferring, communicating or negotiating climate change science to the international policy sphere while maintaining technically correct information. The advantages of the process that were identified by authors are that first, it generates an audience for scientific information; second, it has sufficient checks and balances that maintain the integrity of the science and third, it supports consensus building among international governments.  6.7.2 The Underlying Patterns in the Interviews This investigation into authors' perceptions of the process presents an unanticipated finding. Given the level of science-policy interaction and negotiation in the construction of the SPM and SYR, it was 134  expected that the authors would perceive the document as purely political. What is found through the interviews with authors is that the majority perceived the interaction and negotiation in the review and approval processes as necessary in order to diffuse climate change science into the broader policy sphere. Most interviewees considered the SPM to be the document of governments, accompanied by certain processes and trade-offs, but nonetheless viewed the SPM as a scientifically credible document (see Appendix E).  There are three patterns that emerged from the interviews with the authors. First, the authors all had different understandings of what was happening in the SPM and PRSQ processes. Despite the formal IPCC rhetoric that focussed on translation, many authors saw something different occurring. Given their different interpretations it became clear that each had formulated personal understandings of the process, some to legitimize their scientific involvement, others to be involved in a cross-cutting forum and others still to be included and involved on the inside rather than the outside.  Second, the spectrum of perspectives ranged from the SPM being a credible translation of the scientific findings to it being a politicized document. These opposing ends of the spectrum reflect, respectively, a 'truth speaks to power' and a 'power speaks to truth' model of science-policy. In the former, the author does not view the SPM process as a negotiation but rather as a translation where governments focus on more effective ways of communicating the scientific messages (Interview [40], 2001). This author believes credible scientific findings are being transferred to the policy sphere (Interview [40], 2001). In the latter, the author has a similar view about the role of science in policy; Science is fact that must be used in policy (Interview [43], 2001). This author does recognize the negotiation in the process and perceives this negotiation as governments skewing, manipulating and contaminating the facts (Interview [43], 2001). This conception reverts to a relativistic 'power speaks to truth' model of science in policy where governments use the science for pursue their own positions.  Third, the majority of authors viewed the review and approval processes as the best possible way of ensuring science is integrated into the policy sphere. With one exception, all authors recognized that the SPM and SYR processes were characterized by considerable science-policy interaction. This interaction had varying levels of support, trepidation and in one case resistance among the authors interviewed yet the overall credibility of the process is supported by each of these authors despite identified trade-offs. This suggests that the boundary processes generate a level of confidence among authors in the credibility of the SPM and SYR.  135  6.7.3 The Legitimate and Credible Combination What is argued in this detailed examination of the SPM and SYR innovations is that there is considerable science-policy interaction in their production. The review and approval processes ensure the dual credibility of the information. It is clear that consistency, balance and policy neutrality are not objective criteria but require considerable debate and negotiation. The resulting product is imbued with meaning that is based on underlying assumptions, values and premises used in selection and interpretation. The review and approval processes lead to considerable discussion and debate about acceptable information and its requirements. These boundary processes ensure an attentiveness to alternate meanings associated with the diverse value-frameworks. For instance, the inclusion of both quantitative costs and benefits and issues of fairness (or DES) in the WGIII SPM ensures that information speaks to particular values, frameworks and meanings in the final representation. What informally emerges through these processes is a dialectic arena where scientific representations are interrogated and negotiated by different governmental perspectives.  The IPCC's review process contributes to the bidirectional exchange of materials and discourses between the authors and governments where governments identify areas of inconsistency, imbalance and/or policy prescription. In dealing with the multitudes of governmental comments, authors derive a sense of the different, at times divergent, governmental frameworks and attempt to make revisions that minimize these concerns. A n unanticipated effect of the review process was the way that this type of governmental scrutiny in the initial policy relevant documents influenced the reflexivity among authors. In drafting the ' revised documents, authors were forced to more cautiously consider ways the materials were represented and ways they would be interpreted in the non-scientific governmental sphere.  The approval process was a formalized forum where the final representations of material could be approved line-by-line. What was observed was far more fractured and tense than an approval of the revised scientific message. At this interface there was a complex set of practices that facilitated negotiation over what is and is not a consistent, balanced and policy neutral message. Different governments interpreted policy neutrality differently, and the disentangling of the policy prescriptive from the scientific messages required considerable negotiation. In this forum the values embedded in the science were mediated through diverse interpretations and frameworks of what was considered to be 'policy neutral enough' (or approvable) information. The face-to-face interaction in the approval process ensures that differing interpretations about what is and what is not policy neutral is expressed and attended to within the confines of the organization. The conflict resolution mechanism - the contact group, ensures an informal exchange between scientists and government officials where logics and rationales for particular interpretations can be made, discussed and debated until agreement on an 136  acceptable representation is reached. In this interaction, what authors and governments can say becomes constrained, with authors attempting to focus on consistent and balanced representations and governments determining the balance and neutrality of those representations. Through these formalized science-policy interactions there is neither a scientization of policy nor a politicization of science but instead each detail of the policy relevant science is interrogated, negotiated and transparently co-produced into hybrid form. In other words, the SPM and SYR act as boundary objects where information is exchanged and representations are achieved that preserve the identities of both communities involved.  The PRSQ's were an attempt to connect the scientific information to the parallel U N F C C C policy process and to include the user in the front-end framing of policy relevance. The PRSQ's acted as a boundary process that attempted to close the loop on what is policy relevant. Rather than using the idealized concept of a unidirectional transfer of the scientific information to the policy community, the PRSQ's were an attempt to include users in determining what is policy relevant at the front-end. This front-end inclusion is in principle an important way of ensuring relevance in the initial selections for the assessment for what may be interpreted on behalf of urgent and irreversible policy issues. Framing is particularly important because as was discussed in Chapter 2, the questions asked, methodologies used and interpretations made in the framing of scientific inquiry influence what tacit assumptions and commitments are made.  In practice, an unanticipated finding with regard to the PRSQ's related to whether these questions were a genuine attempt to increase policy relevance and interdisciplinary synthesis to further connect to the COP policy process. That these questions were drafted by a few nations, rather than being deliberated and debated in the international policy sphere, minimized the front-end involvement of governments in determining and fashioning questions with regard to what is policy relevant. Although the IPCC has taken initial steps toward, what is referred to here as strong co-production, with the introduction of the PRSQ's, it was not achieved.  84  This was due to the fact that the questions were framed after the  assessment of literature was already complete rather than being framed through a deliberative process among the diverse 'user' community. For these reasons the extent to which policy relevance influenced the selection and interpretation of information was minimized and limited to authors' anticipation of what is and is not policy relevant. The inclusion of users in framing the problem and the informational requirements to address it, make transparent the implicit credibility cycle between science and policy  A number of scientists expressed dismay stating that the PRSQ's were poorlyframedand uninteresting. It is unclear how policymakers perceived them. One governmental official describes the SYR as "internally consistent but anti-climatic" (Interview [29]) and questions the purpose and use of the SYR. He states that due to the fact that "you cannot re-approach a topic in the SPM [because of the Simple Rule] it is necessary to think more deeply about the [policy relevant scientific] questions" (Interview [29]). 84  137  where scientists pursue instrumental gaps in policy relevant research and in these efforts are more fundable and supportable by the policy community  85  6.8 Out with the Idealized Model of Science-Policy, In with a Pragmatic Model This chapter has argued that scientific information is not comprised of neutral 'facts' but has underlying assumptions and commitments embedded at every stage of knowledge production. In the mandated climate assessment in the IPCC, these assumptions and commitments are transferred into a non-scientific sphere and are therefore used to support differing values and goals. Different frameworks are used to judge what is considered consistent, balanced and policy neutral scientific material; these criteria do not arise from the scientific information itself but are negotiated at the interface with the policy community. The national governments involved in the IPCC can have divergent frameworks for determining what is and what is not balanced and neutral. The definitions of what is acceptable and approvable information are reviewed and negotiated in a transparent forum.  Through ad hoc attempts to increase the credibility and legitimacy of IPCC documents to the international policy sphere, the IPCC has indirectly modified the science-policy relationship. A new model has replaced the unidirectional model of science. What are the appropriate ways of viewing it? The IPCC has developed complex practices and protocols for sophisticated science-policy interaction in its efforts to produce policy relevant and policy neutral scientific information. Discourses are exchanged through the review process where tacit understandings of what is considered acceptable and relevant material are transferred back from the policy community to the scientists. The approval process facilitates a formal negotiation where representations that are technically correct and politically approvable are drafted ensuring international governments and scientists agree on a final interpretation of the science. The SPM and SY/R produce hybrid climate change information that is neither scientific or policy but some combination of both.  These boundary processes introduce normative dimensions into the information. These normative dimensions must be considered when making decisions for the future. Scientific projections are based on contemporary social contexts and analytical models that are projected into the future. Therefore having a technocratic or unidirectional model of science in policy reinforces contemporary structures and knowledge when new structures and new foundations for thinking may be required or desired. The Special Report on Emission Scenarios (SRES) models closes the unidirectional loop in the IPCC whereby  Latour and Woolgar (1979) refer to this as the 'credibility cycle' where scientists produce public meaning through their ability to enrol allies and through the manipulation of resources; this cycle refers to the endless sequence of producing work, receiving recognition, and getting support. 8 5  138  six socio-economic projections are used to determine greenhouse gas emissions in the future based on diverse future development trajectories. The outputs from these models have been introduced into the G C M models in order to develop an understanding of future G H G concentrations, impacts and potential responses. But the general point remains: the socio-political values and norms embodied in those scenarios get carried over into the analysis of the costs and risks associated with different adaptation or mitigation options.  From an empiricist or positivist model that views science as objective and neutral, the transfer of scientific information into the policy community is perceived to contribute to better policy decisions. This model does not work in the IPCC. The selections, interpretations and representations made in the SPM and SYR. are in themselves value-laden. In practical terms, objectivity never implies closure, and there are no facts immune from legitimate dissent. When transferred to the policy sphere, technical information is not unambiguous but instead tacitly transfers normative commitments about what is reasonable, fair, and just. This does not suggest that scientific information cannot offer considerable insight. Recognizing that social and cultural commitments are embedded in technical information highlights the need for technical information used in the policy sphere to be interrogated and negotiated by a broader community. In the case of the IPCC, the 'user' community are international policymakers whose interests are represented through government officials (and to a lesser extent international environmental and industry non-governmental organizations).  86  The potential stakes involved in a  changing climate, and global and regional environmental problems in general, and the uncertainty of the future require normative decisions to be made about how to live in the future.  Through the Imbued Meaning framework, the focus on scientific processes for generating credibility and legitimacy rather than scientific content can foster new ways for conceptualizing the use of science in policy. If empirical information of the use of science in policy or mandated science is taken seriously, what is recognized is that the idealized unidirectional transfer does not occur but instead social networks and users negotiate the value of scientific information for its particular policy context. Therefore developing managed negotiations between scientists and users of science can 1) minimize underlying value conflicts and politicization among users in the international sphere, 2) craft information considered more policy neutral thereby increasing the validity and social robustness of scientific information in the policy sphere, and 3) make the interpenetration of science and policy transparent. A n interactionist model acknowledges the need for scientific analysis in the policy sphere while also recognizing that important  The inclusion of a more diverse user community is necessary to consider in this international forum. For instance, environmental and industry NGO's are involved in the review process but are not included in the approval process. It is interesting to consider what kind of precedent would be set if NGO's were included in this forum and were given veto authority similar to a state. 8 6  139  policy decisions about how to understand climate change and the ways to respond to it are not only technical decisions but are normative decisions as well. Rather than being guided by the tacit values embedded in technical analysis, it becomes important to derive ways of interrogating these assumptions and values in order to conceive of broader options and opportunities for considering how to live in the future. Rather than using a conventional science-policy model to decipher what is occurring in the IPCC, an interactionist model provides a more desirable way of viewing and interpreting construction of policy relevant scientific information in the SPM and SYR than a unidirectional model.  CHAPTER VII: Concluding Remarks and Recommendations In mandated science forums at the international scale, science is often considered to be a reservoir of policy neutral facts that can be used to legitimate policy decisions (Lidskog and Sundqvist 2002). Yet this taken-for-granted role for science in the policy sphere does not address the considerable constructivist research that has examined the use of science at this interface. Nor does it consider the social studies of science (SSS) literature that has accumulated over the past four decades and brings the status of science and its claims under critical purview. The Imbued Meaning framework outlined in this dissertation attempts to integrate these literatures in a pragmatic way to affirm a role for science in decision-making processes in a way that recognizes that science is not unidirectionally transferred into the policy sphere but is negotiated in situ at the interface between science and policy (Jasanoff 1990, 1995; Jasanoff and Wynne 1998). Credibility is dynamically constituted as people and organizations develop new forms of social relationships (Shapin 1994 in Miller 2001, 496). Credibility and legitimacy are not found in either the scientific or the policy domain but are created through collective expression and normative agreement.  The Imbued Meaning framework utilizes a constructivist focus on process rather than content in order to map out a pragmatic approach for designing mandated science processes that enhance the credibility, legitimacy and relevance of information at the science-policy interface. This framework sits between the social studies of science literature and the constructivist science-policy literature. It was derived as a way to address the considerable back and forth that occurs in environmental policy debates between concerns both about the scientization of policy and about the politicization of science. It is inadequate to approach technical policy decisions by waiting for conclusive evidence from the science (what is and is not conclusive can always be challenged especially in dynamic, non-linear and multivalent environmental problems). Instead an innovative and pragmatic approach is proposed; one that manages the interactions between science and policy at the interface in order to co-produce policy relevant hybrid information. The principal conclusions of this research address three main questions, (i) in principle what can a 140  constructivist approach contribute to viewing and examining the issues of the scientization of policy and the politicization of science; (ii) in practice how does this view corroborate with the processes and protocols in the IPCC that facilitate science-policy interaction for the production of policy relevant scientific information that is simultaneously credible and legitimate; and (iii) what are some of the broader implications of this approach for the use of science in policy.  7.1 Constructivist vs. Unidirectional If we step back and examine what the constructivist literature tells us in principle, we find that complex conceptual, social and material processes interpenetrate in areas of scientific inquiry to produce sociotechnical hybrids (Latour 1987, 1993). In combination, constructivist literature in the social studies of science and in science-policy studies suggest that science is imbued with social and cultural assumptions in the construction of scientific claims through empirical and/or experimental inquiry and also through the extrapolations made as scientific claims and assessments move into the socio-political sphere. The social studies of science literature argues that social processes influence the scientific practices of discovery and verification s and that underlying conceptual, social and material processes are interpenetrating forces in observations, data collection and the interpretation of data. The social studies of science argues that in order to develop a realistic rather than idealized or rhetorical understanding of science it is necessary to examine the contextual practices and processes involved in the arena of knowledge production. This understanding of the status of science has considerable implications for the use of science in the policy sphere.  The conventional unidirectional model of science-policy, where science is in principle translated and transferred into the policy sphere, is illustrative of an idealized understanding of science in policy that occurs. Constructivist science-policy scholarship finds that, in practice the value of science in the policy sphere is a result of negotiations over boundary work and the value of particular scientific representations at the interface (Jasanoff 1990, 1995). The suggestion is that:  The use of science as well as the production of scientific data - is always contingent in relation to the social order (Lidskog and Sundqvist 2002). The mandated science literature suggests that the use of science in policy involves hybrid activities that are neither science nor policy. The attempt to draw boundaries and to purify the roles of science and policy in theory is an attempt to unravel and manage information that already exists in mutually reinforcing institutions, which are indistinguishable in practice (Jasanoff and Wynne 1998). Sciencepolicy scholarship shows how science and policy communities and institutions are permeable and  141  dependent on one another (Jasanoff and Wynne 1998; Shackley and Wynne 1996). Attempts to purify the realms of science and politics is said to lead to a proliferation of hybrids:  If we link together in one single picture the work of purification and the work of mediation that gives it meaning, we discover retrospectively, that we have never been truly modern (Latour 1993, 91). The co-production of knowledge occurs whether it is managed or not. The novelty of the Imbued Meaning framework is in its attempt to facilitate and encourage science-policy interaction in a transparent way rather than continuing attempts to disentangle and purify these domains and their interactions.  7.2 The IPCC as Process: Relevant Enough and Neutral Enough The IPCC is a complex adaptive system. There has been an underlying understanding in the developments and adaptations of the IPCC that mandated scientific findings have little practical meaning unless they are valued by the broader socio-political sphere. Over the past three iterations of the IPCC, the Bureau has responded to the criticisms and demands of the broader socio-political sphere. The result has led to complex arrangements, processes and practices that continue to stabilize the IPCC process as both credible and legitimate in the international sphere; identified by the continual sponsorship of governments through the TSU's, the continual participation of credible scientists and governments that play both actor and audience for its policy relevant scientific findings.  Efforts of the IPCC to increase its legitimacy in the governmental sphere have resulted in a series of procedural adaptations over the three iterations of the IPCC assessment reports. Unexpectedly these adaptations have increased government involvement and have simultaneously increased science-policy interaction in the scientific assessment process. The introduction of the intergovernmental mechanism, which included governments in the review of the scientific information, was the initial effort to include governments to ensure the legitimacy of climate change science among international governments and officials. In the SAR, the demarcation of science from policy relevant science through the addition of the ad hoc products of the Summary for Policymakers (SPM) and the Synthesis Report (SYPv) was a response to public challenges questioning the impartiality of the process and the credibility of the IPCC products. The SPM and SYR were introduced as a considerable project of boundary work, to demarcate the . underlying science in the W G reports from the policy relevant science. This was done to inoculate the underlying science assessment from the policy context while at the same time securing an audience for policy relevant findings. The inclusion of governments in translating 'policy relevant but not policy prescriptive scientific information' through the managed SPM and SYR interfaces in the T A R was a practical way to 1) ensure that the underlying assessment reports were one-step removed from the policy 142  sphere and to 2) explicitly include governments in the process of determining the legitimacy of policy relevant scientific findings.  In principle, the SPM and SYR were attempts to encourage scientists to  synthesize and effectively translate policy relevant scientific findings to the international policy sphere. In practice, however, the underlying processes involved in constructing "policy relevant but not policy prescriptive" findings required considerable interrogation, deliberation, and negotiation and fostered significant science-policy interaction.  7.3 Interactive Boundary Processes  In practice, this thesis shows that the SPM and SYR have three boundary processes that manage negotiations at the interface in determining what are acceptable policy relevant findings. The  review  process contributes to governmental legitimacy and indirect exchange of information back to scientists (who are required to manage and address diverse and contradictory comments if they are not demonstrably wrong). This process contributed to learning among veteran authors who anticipated how particular framings and representations of scientific findings would be received in the intergovernmental plenary and who became strategic in framing scientific findings and deciphering what could be said, what should be said and what should not be said in the political forum. The approval process is a formal and managed negotiation between the scientific and governmental communities. The fact that each nation is equipped with veto authority in the line-by-line approval of t