L'objectif de la thèse est de mettre au jour les différents modes de représentation du discours autre et de déterminer leur fonctionnement dans la construction du dessin de presse en tant que commentaire de l'événement médiatique. En tant que genre hybride et humoristique, le dessin de presse emprunte ses caractéristiques rhétoriques, thématiques, discursives et formelles à la caricature, au dessin d'humour, à la bande dessinée et au commentaire journalistique. Quant aux différentes formes d'humour dans le dessin, elles sont liées étroitement à la ligne éditoriale du périodique de publication et elles sont censées créer une connivence avec le lecteur. Le cadre théorique repose sur les recherches d'Authier-Revuz. Elle détermine le lien d'un acte d'énonciation avec les différents modes de représentation du discours autre en s'appuyant sur leurs caractéristiques sémiotiques, syntaxiques et sémantiques spécifiques. Également, la représentation du discours autre relève du dialogisme bakhtinien et le lecteur doit parfois faire appel à sa mémoire discursive, sémantique, historique et collective ainsi qu'à l'implicite pour reconnaître certains énoncés comme relevant de la représentation du discours autre. Le cadre théorique ainsi posé est appliqué à un corpus de 150 dessins de presse publiés dans le quotidien Le Monde, dans l'hebdomadaire Courrier international et exposés au Mémorial de Caen entre 2010 et 2014. Les conclusions des analyses sont les suivantes : premièrement, le discours direct, le discours indirect, la modalisation en assertion seconde et la modalisation autonymique d'emprunt sont les modes de représentation du discours autre présents dans le dessin de presse. Deuxièmement, il apparaît que, malgré le repérage systématique des caractéristiques spécifiques de chaque mode, il est parfois impossible de trancher définitivement sur une forme. Seules les connaissances du lecteur peuvent l'inciter à trancher en faveur d'un mode ou d'un autre. Troisièmement, les modes de représentation du discours autre ainsi déterminés participent de la construction d'un commentaire critique, parfois humoristique, de l'événement médiatique. / The aim of the dissertation is to identify different forms of reported speech and to highlight their importance in the construction of cartoons functioning as news comments in the press. The "cartoon" is a hybrid and humorous genre, which owes its rhetorical, thematic, discursive and formal characteristics to the caricature, the humorous drawing, the comic strip and the press comment. The different types of humor in the cartoon are closely connected to the editorial line of the periodical and are supposed to create a certain complicity between the cartoonist and his reader. The theoretical framework of the dissertation is based on Authier-Revuz's work. She identifies different types of reported speech by their specific semiotic, syntactic and semantic characteristics. Reported speech also falls within dialogism as Bakhtin defined it. Readers need their discursive, semantic, historic and collective memory as well as a certain proficiency in detecting implicit messages in order to understand some occurrences of reported speech. The theoretical framework is put into practice on a corpus of 150 cartoons published in the French daily newspaper Le Monde, in the French weekly review of international press Courrier international or exposed at the "Mémorial de Caen" from 2010 to 2014. One can conclude that firstly, indirect speech, direct speech, "modalisation en assertion seconde" and "modalisation autonymique d'emprunt" are found in the cartoons. Secondly, it can be difficult to attribute a type to an utterance of reported speech even when the analyst systematically checks all utterances for specific characteristics. Only the reader's former knowledge can help him choose one of the forms. Thirdly, the cartoonist uses reported speech to create a critical and sometimes humorous comment on events and their treatment in the media.

Este trabalho propõe analisar a forma como se constitui a escrita de alunos de uma turma de 1º ano do ciclo II do ensino fundamental, no momento em que escrevem para atender a demandas institucionais dadas no contexto do Programa Ler e Escrever, da Secretaria Municipal de Educação de São Paulo. Os 26 textos narrativos que compõem seu corpus foram escritos durante o ano letivo de 2010, enquanto atuávamos como professor regente dessa turma, por esse grupo de estudantes, o qual se caracterizava como recém-alfabetizado, com idades entre a pré-adolescência e a adolescência e com histórico de fracasso escolar. De maneira mais específica, os objetivos de nosso trabalho se voltam para a compreensão dos modos como os alunos registram, em seus textos escritos, duas instâncias de diálogo: a) um diálogo que se estabelece com as condições escolares imediatas de produção da escrita e b) um diálogo que se constroi no processo de formulação do texto narrativo, relacionado aos modos de gestão e transmissão do discurso de outrem ou discurso citado no corpo da narrativa. Essa problemática será abordada com base na contribuição fundadora de Bakhtin (1929/2002) e de estudos do texto e do discurso conforme propostos em Jurado Filho (1993). Do ponto de vista de sua metodologia, levando em conta o modo com o qual abordamos os dados, trata-se de uma pesquisa qualitativa, voltada para a análise textual do discurso narrativo em produções escritas de alunos. Além disso, a fim de empreender uma discussão que nos ajude na compreensão da aprendizagem escrita a partir da realidade pedagógica que analisamos, buscaremos referências nos estudos sobre os conceitos de alfabetização e de letramento (SOARES 1998, 2004; MORTATTI, 2004; COLELLO e SILVA LEITE, 2010) e, também, nos autores que têm se debruçado sobre a questão da escrita na escola a partir de uma perspectiva dialógica da linguagem e da compreensão dos espaços de ensino de língua portuguesa como espaços de interação (GERALDI 1984, 1996, 2003; VAL e ROCHA, 2005). / This research proposes to examine the constitution of writing produced by elementary school students at the 1st year of the 2nd cycle, when writing to meet institutional demands as prescribed by the Programa Ler e Escrever (Reading and Writing Program) of the São Paulo City Department of Education. The 26 narrative texts that constitute our body of research were written during the 2010 school year, while we worked as a classroom teacher of these students, by a group of students characterized as newly literate preadolescents and adolescents with a history of school failure. Specifically, the objective of our work is to understand the ways in which students record, in their written texts, two dialogue instances: a) a dialogue which is established with the immediate conditions of writing production at school, and b) a dialogue built on the process of production of narrative texts, related to the modes of management and transmission of the speech of others or speech cited in the body of the narrative. This issue will be addressed based on the founding contribution of Bakhtin (1929/2002) and on text and discourse studies proposed by Jurado Filho (1993). From the point of view of its methodology, taking into account the way in which data was reviewed, we classify this research as qualitative, focused on textual analysis of narrative discourse in written production of students. Furthermore, in order to engage in a discussion which will help us understand the writing learning process in the pedagogical context we reviewed, references will be sought in the studies on the concepts of literacy and lettering (SOARES 1998, 1994; MORTATTI, 2004; COLELLO and SILVA LEITE, 2010), as well as in the work of authors who have been addressing the issue of school writing based on a dialogical perspective of language and on the understanding of the spaces for teaching Portuguese as interaction spaces (GERALDI 1985, 1996; VAL and ROCHA, 2005).

Nesta pesquisa teve-se como objetivo analisar os arranjos de vozes estabelecidos em textos jornalísticos sobre avaliações externas. De modo mais específico, a proposta foi observar em textos publicados pela Folha de S.Paulo sobre o Ideb (Índice de Desenvolvimento da Educação Básica) e o Pisa (Programa Internacional de Avaliação de Estudantes) os seguintes aspectos: a caracterização dos locutores presentes nos textos, tanto como autores, entrevistados ou fontes jornalísticas; a heterogeneidade enunciativa e formas de inserção do discurso relatado; os modos por meio dos quais as diferentes vozes presentes nos textos são articuladas a fim de sustentar determinadas posições do jornal; e as relações interdiscursivas e dialógicas estabelecidas entre os textos e os elementos materializados no jornal. Para a constituição do corpus, foram selecionados vinte e quatro textos sobre o Ideb 2009 e dezessete sobre o Pisa do mesmo ano, todos publicados pela Folha de S.Paulo em 2010, logo após a divulgação dos resultados. Optou-se pelo uso da versão impressa, disponível no acervo digital do jornal, para que fosse possível examinar, além do conteúdo escrito, a disposição do texto e das imagens no material publicado. Como referencial teórico para as análises, foram mobilizados estudos pertencentes, principalmente, à Análise do Discurso e a diferentes correntes acerca da enunciação. Os conceitos basilares foram dialogismo (BAKHTIN, 1997, 2010), heterogeneidade enunciativa e discurso relatado (AUTHIER-REVUZ, 1990, 1999, 2004), interdiscurso (PÊCHEUX, 1993) e locutor (DUCROT, 1987). Da esfera jornalística, foram abordados também os conceitos de fontes e gêneros jornalísticos com base, sobretudo em Melo (1985) e Lage (2008). Por meio das análises realizadas, concluiu-se que, com relação aos locutores, nos textos publicados pelo jornal, predominam enunciados provenientes da esfera governamental e de think tanks, sendo que a esses últimos é atribuído o papel de fontes expert, aquelas que detêm o conhecimento teórico capaz de debater os resultados divulgados e apresentar propostas de ação. Ao saber acadêmico é relegado papel secundário, tanto quantitativamente quanto pelas maneiras como os enunciados associados a representantes de universidades são dispostos nos textos. No que diz respeito ao conteúdo, há predominantemente uma tendência à polarização entre ensino público e privado, com valorização das práticas do segundo em relação ao primeiro; consideração dos dados obtidos pelas avaliações como rankings e propagação do discurso da competitividade. / This study aims to investigate the arrangement of voices established in journalistic texts on external assessments. More specifically, this study focused on how texts published in the Folha de S.Paulo newspaper on the Ideb (Basic Education Development Index) and the Pisa (Programme for International Student Assessment) addressed the following aspects: characterization of speakers present in the texts, as authors, interviewees and journalistic sources; enunciative heterogeneity and reported speech insertion; how the different voices present in the texts are articulated in order to sustain some of the newspapers positions; and interdiscursive and dialogical relations among texts and the elements materialized in the newspaper. To build the corpus, twenty-four texts on the 2009 Ideb and seventeen on the Pisa of the same year were selected from the Folha de S.Paulo newspaper in 2010. The texts came out shortly after the release of Ideb and Pisa results. The printed texts were obtained from the newspapers digital database in order for the study to address both written content as well as text and image layout. The theoretical framework of choice mobilizes Discourse Analysis and various lines of thought in Enunciation. The fundamental concepts operated with were dialogism (Bakhtin, 1997, 2010), enunciative heterogeneity and reported speech (AUTHIER-REVUZ, 1990, 1999, 2004), interdiscourse (PÊCHEUX, 1993) and enunciator (DUCROT, 1987). From journalism, the concepts of sources and journalistic genres based mainly on Melo (1985) and Lage (2008) were approached. It was found that, with respect to the enunciators present in the Folha de S.Paulo texts, statements made by government authorities and think tanks prevailed. The latter is assigned the role of expert source, one holding theoretical knowledge capable of discussing the disclosed results and making proposals for action. Academic knowledge is relegated to a secondary role, not only quantitatively but also in the way statements associated with university enunciators are arranged in the texts. With regard to content, the prevailing trend is toward polarizing public and private education, thus valuing the practices of the latter in detriment to the former; translating data obtained from assessments as ranking lists, and promoting the discourse of competition.

Nesta pesquisa teve-se como objetivo analisar os arranjos de vozes estabelecidos em textos jornalísticos sobre avaliações externas. De modo mais específico, a proposta foi observar em textos publicados pela Folha de S.Paulo sobre o Ideb (Índice de Desenvolvimento da Educação Básica) e o Pisa (Programa Internacional de Avaliação de Estudantes) os seguintes aspectos: a caracterização dos locutores presentes nos textos, tanto como autores, entrevistados ou fontes jornalísticas; a heterogeneidade enunciativa e formas de inserção do discurso relatado; os modos por meio dos quais as diferentes vozes presentes nos textos são articuladas a fim de sustentar determinadas posições do jornal; e as relações interdiscursivas e dialógicas estabelecidas entre os textos e os elementos materializados no jornal. Para a constituição do corpus, foram selecionados vinte e quatro textos sobre o Ideb 2009 e dezessete sobre o Pisa do mesmo ano, todos publicados pela Folha de S.Paulo em 2010, logo após a divulgação dos resultados. Optou-se pelo uso da versão impressa, disponível no acervo digital do jornal, para que fosse possível examinar, além do conteúdo escrito, a disposição do texto e das imagens no material publicado. Como referencial teórico para as análises, foram mobilizados estudos pertencentes, principalmente, à Análise do Discurso e a diferentes correntes acerca da enunciação. Os conceitos basilares foram dialogismo (BAKHTIN, 1997, 2010), heterogeneidade enunciativa e discurso relatado (AUTHIER-REVUZ, 1990, 1999, 2004), interdiscurso (PÊCHEUX, 1993) e locutor (DUCROT, 1987). Da esfera jornalística, foram abordados também os conceitos de fontes e gêneros jornalísticos com base, sobretudo em Melo (1985) e Lage (2008). Por meio das análises realizadas, concluiu-se que, com relação aos locutores, nos textos publicados pelo jornal, predominam enunciados provenientes da esfera governamental e de think tanks, sendo que a esses últimos é atribuído o papel de fontes expert, aquelas que detêm o conhecimento teórico capaz de debater os resultados divulgados e apresentar propostas de ação. Ao saber acadêmico é relegado papel secundário, tanto quantitativamente quanto pelas maneiras como os enunciados associados a representantes de universidades são dispostos nos textos. No que diz respeito ao conteúdo, há predominantemente uma tendência à polarização entre ensino público e privado, com valorização das práticas do segundo em relação ao primeiro; consideração dos dados obtidos pelas avaliações como rankings e propagação do discurso da competitividade. / This study aims to investigate the arrangement of voices established in journalistic texts on external assessments. More specifically, this study focused on how texts published in the Folha de S.Paulo newspaper on the Ideb (Basic Education Development Index) and the Pisa (Programme for International Student Assessment) addressed the following aspects: characterization of speakers present in the texts, as authors, interviewees and journalistic sources; enunciative heterogeneity and reported speech insertion; how the different voices present in the texts are articulated in order to sustain some of the newspapers positions; and interdiscursive and dialogical relations among texts and the elements materialized in the newspaper. To build the corpus, twenty-four texts on the 2009 Ideb and seventeen on the Pisa of the same year were selected from the Folha de S.Paulo newspaper in 2010. The texts came out shortly after the release of Ideb and Pisa results. The printed texts were obtained from the newspapers digital database in order for the study to address both written content as well as text and image layout. The theoretical framework of choice mobilizes Discourse Analysis and various lines of thought in Enunciation. The fundamental concepts operated with were dialogism (Bakhtin, 1997, 2010), enunciative heterogeneity and reported speech (AUTHIER-REVUZ, 1990, 1999, 2004), interdiscourse (PÊCHEUX, 1993) and enunciator (DUCROT, 1987). From journalism, the concepts of sources and journalistic genres based mainly on Melo (1985) and Lage (2008) were approached. It was found that, with respect to the enunciators present in the Folha de S.Paulo texts, statements made by government authorities and think tanks prevailed. The latter is assigned the role of expert source, one holding theoretical knowledge capable of discussing the disclosed results and making proposals for action. Academic knowledge is relegated to a secondary role, not only quantitatively but also in the way statements associated with university enunciators are arranged in the texts. With regard to content, the prevailing trend is toward polarizing public and private education, thus valuing the practices of the latter in detriment to the former; translating data obtained from assessments as ranking lists, and promoting the discourse of competition.

This thesis features a qualitative study of student source use and speech representation in two corpora of review essays that acknowledges the complexity of classroom writing contexts and the rhetorical nature of school genres. It asks how students engage with the texts they review, for what reasons, and in response to what aspects of the writing context. When considered as a distinct genre of student assignment, review essays make for a particularly interesting study of source engagement because they challenge students to maintain an authoritative voice as novices evaluating the work of an expert. In addition, citation issues in the review assignment might not be as obvious to students or their instructors as they would be, for instance, in a research paper for which multiple sources are consulted and synthesized. The review essays interrogated in this study were collected with appropriate ethics clearance from two undergraduate history courses. The analysis is extended to a small corpus of published reviews assigned as model texts in one of these courses. The study features a robust method that combined applied linguistics and discourse analysis to tease out connections between the grammatical structures of speech reports and their argumentative roles. This method involved a recursive process of classifying speech reports using Swales’ (1990) concepts of integral and non-integral citation, Thompson and Yiyun’s (1991) classifications of speech act verbs, and Vološinov (1929/1973) and Semino and Short’s (2004) models of speech reporting forms. In addition, the analysis considered the influence of the writing context on the students’ citation practices and took into account theories of rhetorical genre and student identity. The results show connections between assignment instructions and the effective and problematic ways students engaged with the texts they reviewed, such as a correlation between a directive to reduce redundancy and the absence of in-text attributions. Most notably, this study offers a fluid set of descriptors of the forms and functions of speech reports in student coursework that can be used by students, educators, plagiarism adjudicators, as well as scholars of rhetoric and composition, to illuminate some of the methods and motives of student source use.

This thesis is an investigation of talk in a therapeutic setting. It takes discursive psychology as the main influence theoretically, and also draws on the rigorous analytical techniques of conversation analysis (CA). The data was collected in various family therapy settings in the U.K., both residential and non-residential videotapes made during those sessions These recordings were made by therapists for their own use initially, and were not produced especially for this project. Videotapes were transcribed according to standard CA conventions, and subsequently analysed. One of the primary research questions has been to examine empirically mental state language as used in the therapeutic setting. Secondly, it has been to examine accounting practices and the production of versions of events as 'fact'. Thirdly, the aim has been to consider the practical implications of asymmetry as a participants' concern. As a unifying and over-arching analytic interest the use of reported speech in each of these other aspects has been investigated to assess its role in their production. The conclusions of the thesis demonstrate that participants themselves orient to one another's minds as accessible and reportable entities, and that speech is treated as reflective of inner thought. Furthermore, where speech is reported in the therapeutic setting, it is frequently used to validate and to evidence claims about other people's 'psyche'.

This thesis features a qualitative study of student source use and speech representation in two corpora of review essays that acknowledges the complexity of classroom writing contexts and the rhetorical nature of school genres. It asks how students engage with the texts they review, for what reasons, and in response to what aspects of the writing context. When considered as a distinct genre of student assignment, review essays make for a particularly interesting study of source engagement because they challenge students to maintain an authoritative voice as novices evaluating the work of an expert. In addition, citation issues in the review assignment might not be as obvious to students or their instructors as they would be, for instance, in a research paper for which multiple sources are consulted and synthesized. The review essays interrogated in this study were collected with appropriate ethics clearance from two undergraduate history courses. The analysis is extended to a small corpus of published reviews assigned as model texts in one of these courses. The study features a robust method that combined applied linguistics and discourse analysis to tease out connections between the grammatical structures of speech reports and their argumentative roles. This method involved a recursive process of classifying speech reports using Swales’ (1990) concepts of integral and non-integral citation, Thompson and Yiyun’s (1991) classifications of speech act verbs, and Vološinov (1929/1973) and Semino and Short’s (2004) models of speech reporting forms. In addition, the analysis considered the influence of the writing context on the students’ citation practices and took into account theories of rhetorical genre and student identity. The results show connections between assignment instructions and the effective and problematic ways students engaged with the texts they reviewed, such as a correlation between a directive to reduce redundancy and the absence of in-text attributions. Most notably, this study offers a fluid set of descriptors of the forms and functions of speech reports in student coursework that can be used by students, educators, plagiarism adjudicators, as well as scholars of rhetoric and composition, to illuminate some of the methods and motives of student source use.

Ce travail se propose d’étudier la postposition totale ou partielle de l'énoncé rapportant dans le discours rapporté direct et indirect. L’approche grammaticale, dite "classique" du discours rapporté (DR) en général et des incises en particulier se heurte à l’extrême diversité des données empiriques à ce sujet. Envisager les incises sous l’angle formel d’un phénomène grammatical, c’est à la fois en dessiner les contours, mais également poser les limites inhérentes à toute forme. Confrontés à la réalité des incises, les contours rassurants de la forme se distendent jusqu’à en donner une image déformée voire paradoxale sous certains aspects, bien loin des clichés véhiculés par l’emploi métalinguistique de ce terme en français. Il s’agit de s’intéresser au fonctionnement spécifique des incises de DR en anglais, de réhabiliter, en quelque sorte, la position incise dont elles tirent leur nom. Cette réhabilitation de nature textuelle, fondée sur l’étude de leur métamorphose au fil des textes, permettrait d'envisager les incises comme un lieu de stratégie narrative plus qu'un élément de rupture énonciative. / This thesis investigates the use of reporting clauses and reporting parentheticals in final and medial position in direct and indirect (reported) speech. Traditional grammatical models seem to offer no satisfactory explanation as regards to final and medial positions of the reporting clause or parenthetical within the sentence. Though grammatical descriptions give formal structure to the phenomenon, they also tend to limit its scope and need to be stretched to accommodate the reality of reporting clauses and parentheticals in literary texts.Analysing medial and final reporting clauses within the framework of French théorie de l’énonciation shows that they blur the frontier between reported and non-reported speech more often than not. From a literary point of view, medial and final positions can be considered as choices made for stylistic reasons, in which case the emphasis is put on reporting clauses as creating textual cohesion rather than causing rupture within the narrative.

Este trabalho propõe analisar a forma como se constitui a escrita de alunos de uma turma de 1º ano do ciclo II do ensino fundamental, no momento em que escrevem para atender a demandas institucionais dadas no contexto do Programa Ler e Escrever, da Secretaria Municipal de Educação de São Paulo. Os 26 textos narrativos que compõem seu corpus foram escritos durante o ano letivo de 2010, enquanto atuávamos como professor regente dessa turma, por esse grupo de estudantes, o qual se caracterizava como recém-alfabetizado, com idades entre a pré-adolescência e a adolescência e com histórico de fracasso escolar. De maneira mais específica, os objetivos de nosso trabalho se voltam para a compreensão dos modos como os alunos registram, em seus textos escritos, duas instâncias de diálogo: a) um diálogo que se estabelece com as condições escolares imediatas de produção da escrita e b) um diálogo que se constroi no processo de formulação do texto narrativo, relacionado aos modos de gestão e transmissão do discurso de outrem ou discurso citado no corpo da narrativa. Essa problemática será abordada com base na contribuição fundadora de Bakhtin (1929/2002) e de estudos do texto e do discurso conforme propostos em Jurado Filho (1993). Do ponto de vista de sua metodologia, levando em conta o modo com o qual abordamos os dados, trata-se de uma pesquisa qualitativa, voltada para a análise textual do discurso narrativo em produções escritas de alunos. Além disso, a fim de empreender uma discussão que nos ajude na compreensão da aprendizagem escrita a partir da realidade pedagógica que analisamos, buscaremos referências nos estudos sobre os conceitos de alfabetização e de letramento (SOARES 1998, 2004; MORTATTI, 2004; COLELLO e SILVA LEITE, 2010) e, também, nos autores que têm se debruçado sobre a questão da escrita na escola a partir de uma perspectiva dialógica da linguagem e da compreensão dos espaços de ensino de língua portuguesa como espaços de interação (GERALDI 1984, 1996, 2003; VAL e ROCHA, 2005). / This research proposes to examine the constitution of writing produced by elementary school students at the 1st year of the 2nd cycle, when writing to meet institutional demands as prescribed by the Programa Ler e Escrever (Reading and Writing Program) of the São Paulo City Department of Education. The 26 narrative texts that constitute our body of research were written during the 2010 school year, while we worked as a classroom teacher of these students, by a group of students characterized as newly literate preadolescents and adolescents with a history of school failure. Specifically, the objective of our work is to understand the ways in which students record, in their written texts, two dialogue instances: a) a dialogue which is established with the immediate conditions of writing production at school, and b) a dialogue built on the process of production of narrative texts, related to the modes of management and transmission of the speech of others or speech cited in the body of the narrative. This issue will be addressed based on the founding contribution of Bakhtin (1929/2002) and on text and discourse studies proposed by Jurado Filho (1993). From the point of view of its methodology, taking into account the way in which data was reviewed, we classify this research as qualitative, focused on textual analysis of narrative discourse in written production of students. Furthermore, in order to engage in a discussion which will help us understand the writing learning process in the pedagogical context we reviewed, references will be sought in the studies on the concepts of literacy and lettering (SOARES 1998, 1994; MORTATTI, 2004; COLELLO and SILVA LEITE, 2010), as well as in the work of authors who have been addressing the issue of school writing based on a dialogical perspective of language and on the understanding of the spaces for teaching Portuguese as interaction spaces (GERALDI 1985, 1996; VAL and ROCHA, 2005).

Orientador: Sirio Possenti / Dissertação (mestrado) - Universidade Estadual de Campinas, Instituto de Estudos da Linguagem / Made available in DSpace on 2018-08-06T20:50:09Z (GMT). No. of bitstreams: 1 Fossey_MarcelaFranco_M.pdf: 700627 bytes, checksum: b4e805cd5f2cc0440e63767bdb0513c0 (MD5) Previous issue date: 2006 / Resumo: O trabalho que se segue visa caracterizar, a partir de uma perspectiva discursiva, dois modos distintos de divulgar ciência para leigos em duas revistas de divulgação científica: a Superinteressante e a Pesquisa Fapesp. Partindo de um conhecimento prévio em relação ao público-alvo de cada uma dessas publicações ¿ um público predominantemente jovem e leigo, no primeiro caso, e um público familiarizado com temas de cunho científico, no segundo ¿ buscamos apurar como esses leitores tomam forma a partir de indícios textuais e discursivos. O pressuposto inicial, que tem por base os preceitos teóricos da Análise do Discurso francesa, e mais especificamente a noção de semântica global, tal como proposta por Maingueneau (1984), foi de que, por meio de uma análise de indícios da superfície textual ¿ como o léxico característico e as formas de discurso relatado preferenciais ¿ é possível identificar um suposto leitor, assim como um divulgador e uma imagem de ciência específicos. Isso porque entendemos que discurso e texto são mutuamente dependentes e que, portanto, a escolhas de certos elementos textuais e não de outros são, em boa medida, escolhas discursivamente fundadas ou condicionadas. São, desta perspectiva, indícios de um conjunto de regras que delimita o dizível de cada publicação. Isto é, para além de uma pressuposição editorial por parte das publicações analisadas, procuramos apontar como uma imagem de leitor, de divulgador e de ciência emerge das possibilidades textuais e discursivas dessas revistas. Para tanto, concentramos nossas análises nas reportagens ¿ optamos por selecionar textos cuja temática está centrada em pesquisas realizadas pelas ciências duras ¿ recorrendo a outros gêneros apenas como uma forma de corroborar as hipóteses levantadas no decorrer do trabalho / Abstract: This work aims to characterize, from a discursive approach, two distinct manners of divulging science for general public in two scientific vulgarization magazines: Superinteressante and Pesquisa Fapesp. Taking into account the previous knowledge of who is the target public of each of these magazines ¿ the young layperson, in the first case, and someone closer to the scientific field, in the second case ¿ we describe how this reading public gets ¿materialized¿ through textual and discursive signs. We assumed, based on the French Discourse Analysis, specifically on the notion of global semantic, as proposed by Maingueneau (1984), that by analyzing signs on the textual surface ¿ typical vocabulary or preferred forms of the reported speech, for instance ¿ it is possible to identify the presumed reader, as well as the image of journalist and science emerged from each of these publications. This is a consequence of the understanding that discourse and text are mutually related and, therefore, the use of some textual elements instead of others are, basically, discursive conditioned choices. These choices, from this point of view, point out to the set of discursive rules that circumscribes the ¿speechble¿ of each magazine. This means that, looking beyond any editorial supposition, we tried to demonstrate how it is possible to picture specific images for the reader, the journalist and the science given the textual and discursive possibilities of each of those magazines. For this investigation, we concentrate our analysis on the genre report ¿ particularly texts regarding researches done by ¿hard sciences¿. Other genres are addressed only in order to confirm the hypotheses proposed during the development of this work / Mestrado / Linguistica / Mestre em Linguística