L’expertise scientifique joue un rôle croissant dans la fabrique du politique dans les démocraties. Elle contribue à façonner et orienter les problèmes publics et leurs solutions, de façon relativement autonome par rapport à l’élection des décideurs. Ce rôle est justifié par la complexité grandissante des problèmes et solutions dans les sociétés techniques. Dans cette thèse, nous investiguons les fonctions de l’expertise scientifique jugée crédible par les décideurs. A contrario des prétentions universalistes de la Knowledge Literature et de l’Evidence-Based Policy, nous construisons une théorie de moyenne portée des usages spécifiques, d’une forme d’expertise particulière, dans un contexte donné. Un tel triptyque permet de réconcilier deux approches contradictoires de l’usage de l’expertise. L’une « externaliste » promue par l’analyse des politiques publiques explique son utilisation par le contexte. L’autre « internaliste » défendue par une variété de courants, la justifie par les caractéristiques propres de l’expertise. Pour ce faire, nous investiguons l’utilisation que font en France dans les années 2000 les ministères de la santé et de l’agriculture des expertises collectives INSERM et INRA – méthode d’expertise proche des revues systématiques de la littérature. Dans cette perspective, nous objectivons successivement : les critères d’appréciation des expertises collectives tels qu’utilisés par les décideurs, les types d’usages qu’ils attendent et font de ces expertises, les contextes dans lesquels ils les mobilisent. Notre investigation est aussi l’occasion de clarifier les indicateurs des types d’usage en même temps que les facteurs contextuels les favorisant. Nous mettons en lumière les effets d’autorité et de vérité des expertises collectives sur les acteurs du débat public et de cadrage sur les problèmes publics et leurs solutions. Nous rendons aussi compte de la façon dont l’expertise gagne sa crédibilité scientifique (et son producteur une légitimité scientifique ou réputation) auprès des décideurs. Croisant les travaux de la sociologie des sciences, et des études de la communication, nous établissons que la crédibilité de l’expertise dépend de la conjugaison d’investissements de forme et de fond. Nous montrons que ces dispositifs rhétoriques et pratiques ne sont pas librement accessibles mais exigent des connaissances et compétences particulières. Renouant avec la sociologie de l’expertise, nous constatons que la nature et la réputation de la méthode et des instituts sont le produit d’une histoire longue dans laquelle les chercheurs – leurs positions et représentations, leurs réflexions, actions et interactions – jouent un rôle clef. Sur un plan empirique, notre thèse constitue une des pièces du puzzle de l’émergence des expertises collectives et revues systématiques de littérature en France au cours des vingt dernières années.D’un point de vue méthodologique, notre thèse mobilise une variété de méthodes : qualitative (entretiens individuels, observations ethnographiques, travail d’archive, études de cas) et quantitative (statistiques descriptives, analyse de correspondances multiples et classification hiérarchique ascendante). / Scientific expertise plays a growing role in the policymaking process in democracies. It shapes public problems and their solutions relatively autonomously from the election process. This role is explained by the increasing complexity of problems and solutions in technical societies.In this thesis, we investigate the function of scientific expertise that is deemed credible by policymakers. Unlike scholars in the Knowledge Literature and the Evidence-Based Policy movement, we do not intend to express universalist claims. Instead, we build a middle range theory: one of a type of use of a specific kind of expertise within a particular context. This triptych (type of use-type of expertise-type of context) permits us to reconcile two contradictory approaches of the use of expertise: one “externalist” explains the type of use of expertise by the particular context in which it is used, another “internalist” explains the type of the use by the intrinsic qualities of the specific kind of expertise.In order to do this, we investigate how policymakers within the French ministries for Health and Agriculture use INSERM and INRA collective expertise – an expertise method close to systematic literature reviews.We successively display: the specific criteria policymakers use to assess the credibility of expertise, their intended and effective use of such knowledge, the context in which they order an experts’ report. Our investigation allows us to test and organize a set of indicators and contextual factors of the type of use.We demonstrate that these INSERM and INRA collective expertise appear as an authoritarian and truthful discourse to actors involved in the public debate. They frame public problems and their solutions.We also explain how expertise gains its scientific credibility in the eyes of policymakers. On the basis of recent works in the fields of science and technology studies and communication studies fields, we establish that the credibility of INSERM and INRA collective expertise derives from the implementation of rhetorical and practical devices in their production. We show that such devices are not ready-to-use: their implementation requires a set of knowledge and know-how.We also demonstrate that the very nature and reputation of the expertise and its producers are the result of a long process in which researchers play a key role: their career, position, representation, thoughts, actions and interactions.This thesis constitutes a missing piece in the systematic exploration of the growth of collective expertise and systematic literature reviews in France in the last twenty years.It is based on a variety of qualitative methods (interviews, observations, archive, case studies) as well as quantitative methods (statistics descriptive, factor analysis, hierarchical cluster analysis).

Submitted by Priscilla Araujo (priscilla@ibict.br) on 2016-08-08T19:12:29Z No. of bitstreams: 2 license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) fabiolapinudo-mestrado-2013.pdf: 1403470 bytes, checksum: 8573b6477faa73f521e25676909b3fec (MD5) / Made available in DSpace on 2016-08-08T19:12:29Z (GMT). No. of bitstreams: 2 license_rdf: 0 bytes, checksum: d41d8cd98f00b204e9800998ecf8427e (MD5) fabiolapinudo-mestrado-2013.pdf: 1403470 bytes, checksum: 8573b6477faa73f521e25676909b3fec (MD5) Previous issue date: 2013-12-18 / Esta dissertação abrange a população de pesquisadores da UFRJ, a saber, mestrandos, doutorandos, pós-doutorandos e professores, e investiga seus padrões de uso do serviço de acesso remoto às publicações científicas e ao Portal CAPES através da rede da universidade. Através desse panorama, o presente estudo indaga sobre a existência da relação entre o uso desse acesso, em especial no que se refere ao Portal CAPES, o desempenho na ciência e a busca por credibilidade entre os pesquisadores. O conceito de credibilidade proposto é considerado a partir das reflexões de Bruno Latour e Steve Woolgar (2008) de que ela seria a causa e o resultado do processo de construção e validação dos fatos científicos. Neste estudo, partimos da premissa de que o serviço de acesso remoto às publicações científicas e ao Portal CAPES se constitui em um mecanismo estratégico nesse processo, mas que não é unânime entre os usuários e campos científicos. Este estudo de caso foi estruturado em dois distintos momentos. O primeiro se dá na caracterização geral dos usuários do serviço de acesso remoto com base no tratamento estatístico e interpretação de dados disponíveis no Sistema de Bibliotecas e Informação da UFRJ (SiBI), departamento que administra o serviço. O segundo se dá na identificação e investigação dos motivos para solicitação do acesso remoto e nos diferentes usos que os pesquisadores inscritos no serviço em 2012 fazem do Portal CAPES. Considerando a literatura e as relações estabelecidas no que podemos chamar de “desempenho na ciência”, ou seja, a capacidade que um pesquisador tem de produzir novas informações a partir da publicação de artigos e de seu investimento de tempo nas pesquisas, definimos variáveis e o cruzamento destas para responder à nossa questão de pesquisa. / This dissertation comprehends the population of researchers at UFRJ, i.e., graduate students, postdoctoral fellows and professors, and investigates their use patterns of the remote access service to scientific publications and the CAPES web portal through the computer network of the university. Within this framework, this study questions the existence of the relation between the use of this access, especially vis-à-vis the CAPES portal, scientific performance and the search for credibility among researchers. The concept of credibility we will take into consideration is the one proposed by Bruno Latour and Steve Woolgar (2008), who suggested that it would be the cause and result of the construction and validation process of scientific facts. We will start from the premise that the remote access service to scientific publications and the CAPES portal consists of a strategic mechanism in this process, albeit not unanimous among users and scientific fields. This case study was structured upon two distinct moments. The first one regards the general characteristics of the remote access service users, based on statistical treatment and data interpretation available in UFRJ's Information and Library System (SiBI), the department responsible for the service. The second one regards the identification and investigation of the reasons for soliciting remote access and the various uses researchers subscribing the service in 2012 make of the CAPES portal. Considering the literature and the relations established in what we call “performance in science”, i.e., the capacity a researcher possesses of producing new information out of article publishing and time investment in research, we defined variables and crossed these variables to answer our research question.

Submitted by Maicon Juliano Schmidt (maicons) on 2015-07-09T16:58:59Z No. of bitstreams: 1 Maurício Fernandes da Silva.pdf: 804511 bytes, checksum: 90118227b23715761e729eed7d0e3f2c (MD5) / Made available in DSpace on 2015-07-09T16:58:59Z (GMT). No. of bitstreams: 1 Maurício Fernandes da Silva.pdf: 804511 bytes, checksum: 90118227b23715761e729eed7d0e3f2c (MD5) Previous issue date: 2012-09-25 / Nenhuma / As informações decorrentes da ciência comunicam com o sistema do direito através de perícias judiciais, testemunhos e pareceres, fundamentando a ocorrência e a dimensão do dano ambiental. Constitui-se de extrema relevância a prova científica no âmbito do processo judicial para apuração de danos ambientais, proporcionando ao julgador informações necessárias para expressar sua decisão. No âmbito do direito ambiental, a análise da prova científica pelos Tribunais se dá através de um acoplamento entre o direito e a ciência, segundo a Teoria dos Sistemas. Tal operação exige a formação de critérios previamente definidos. A adequada intersecção destes dois sistemas é necessária para a consolidação da certeza científica frente à necessidade da validade jurídica no processo jurisdicional de tomada de decisão. O sistema jurídico, onde no centro encontram-se os tribunais, atua como uma espécie de filtro de equacionamento do conhecimento científico à sociedade. A definição de critérios para a interpretação jurídica da ciência deve considerar preceitos substanciais e procedimentais. Deve ser preservada a autoridade da ciência, mas incumbe ao direito a competência de ponderar como e de que forma os conceitos científicos serão considerados válidos no processo jurisdicional de tomada de decisão. / The information from the science communicate with the system of law through judicial expertise, testimony and opinions, giving reasons for the occurrence and extent of environmental damage. It consists of highly relevant scientific evidence in legal proceedings for determination of environmental damage, providing the information needed to judge expressing his decision. Under environmental law, the analysis of scientific evidence by the courts is through a coupling between law and science, according to Systems Theory. This operation requires the formation of pre-defined criteria. Adequate intersection of these two systems is necessary for the consolidation of scientific certainty due to the necessity of legal validity in the process of judicial decision making. The legal system where, in the center, are the courts, acts as a kind of filter for solving the scientific knowledge to society. The definition of criteria for the legal interpretation of science should consider substantive and procedural provisions. Should be preserved the authority of science, but it is for the right to consider how and expertise of how the scientific concepts will be considered valid in the process of judicial decision making.