Conselho Nacional de Desenvolvimento CientÃfico e TecnolÃgico / A evidencialidade à uma categoria lingÃÃstica utilizada como estratÃgia que permite a manipulaÃÃo das informaÃÃes quanto à fonte do conhecimento e ao grau de (des)comprometimento do produtor com essa fonte. No intuito de investigar a manifestaÃÃo dessa categoria na construÃÃo das dissertaÃÃes acadÃmicas no portuguÃs brasileiro contemporÃneo, objetiva-se: a) rediscutir o estatuto da evidencialidade como categoria lingÃÃstica; b)analisar as relaÃÃes entre modalidade e evidencialidade; c) verificar as estratÃgias utilizadas no uso das marcas de evidencialidade no trabalho cientÃfico de grau, do tipo dissertaÃÃo; d) identificar, descrever e analisar as marcas lingÃÃsticas evidenciais no trabalho cientÃfico de grau do tipo dissertaÃÃo; e) relacionar o uso de evidenciais com estratÃgias para efeito de (des)comprometimento na construÃÃo textual. Ao identificar e interpretar as marcas evidenciais na construÃÃo dos textos acadÃmicos, a pesquisa busca contribuir para a explicitaÃÃo dos efeitos de sentido vinculados à veiculaÃÃo das informaÃÃes de forma estratÃgica, jà que essas marcas sÃo utilizadas com propÃsitos diversificados, como por exemplo: recurso ao chamado âargumento de autoridadeâ, atenuaÃÃo da responsabilidade em relaÃÃo ao que à dito, modalizaÃÃo no contÃnuo entre a certeza e a nÃo-certeza, sinalizando que algo nÃo està sendo dito de forma categÃrica e sugerindo um grau de (des)comprometimento em relaÃÃo à verdade da proposiÃÃo, como tambÃm um posicionamento crÃtico em relaÃÃo à fonte da informaÃÃo, permitindo uma correta avaliaÃÃo do conteÃdo assimilado pelo leitor, dentre outros. A anÃlise, orientada por pressupostos funcionalistas, conta com uma dimensÃo teÃrica, voltada para a rediscussÃo dos limites conceituais entre as categorias modalidade e evidencialidade, e uma dimensÃo analÃtica, que, em constante diÃlogo com a teoria, investiga, qualitativa e quantitativamente, o uso de marcas da evidencialidade na dissertaÃÃo acadÃmica, trabalho que à requisito para a obtenÃÃo do tÃtulo de mestre. A escolha do gÃnero para a constituiÃÃo do corpus justifica-se pela suposiÃÃo de que esse à um dos gÃneros textuais que apresenta maior quantidade de informaÃÃo cuja fonte nÃo à o prÃprio autor, o que condiciona o uso das marcas de evidencialidade na relaÃÃo observÃvel com o grau de (des)comprometimento do produtor relativamente à informaÃÃo veiculada. A obtenÃÃo das 290 ocorrÃncias para esta pesquisa concretizou-se a partir da organizaÃÃo de um corpus constituÃdo por dez dissertaÃÃes acadÃmicas da Ãrea de CiÃncias Humanas, das quais se utilizaram, como material de anÃlise, a introduÃÃo e a conclusÃo. Os resultados obtidos verificaram os efeitos de sentido associados ao uso de meios de expressÃo evidencial na construÃÃo da argumentaÃÃo em dissertaÃÃes, comprovando que a evidencialidade no discurso acadÃmico, em particular no corpus coletado, à mais utilizada para promover um baixo comprometimento, tendo o verbo como o meio lingÃÃstico mais utilizado e as noÃÃes evidenciais inferencial e citativa como as mais freqÃentes. / Evidentiality is a linguistic category used as a strategy to deal with information, considering the producerâs source of knowledge and his extent of (un)commitment to this source. In order to investigate the manifestation of this category in the construction of academic dissertations written in contemporary Brazilian Portuguese, this work aims to: a) rediscuss the status of the evidentiality as a linguistic category; b) analyse the relationship between modality and evidentiality; c) verify the strategies used to indicate evdentiality in scientific works such as dissertations; d) identify, describe and analyse linguistic marks of evidentiality in such works; e) relate the use of evidentials to strategies of (un)commitment in the production of a text. Though the identification of evidentiality marks in academic texts, this work aims to contribute to explicit the sense effects related to the strategic transmition of information, since these marks are used with different purposes, like: use of the authority argument, reduction of responsibility in relation to what is being said; certainty/uncertainty modalization, suggesting that something is being said in a non-categorical way, suggesting a certain degree of (un)commitment in relation to the truth of the proposition, as well as showing a critical position in relation to the source of information, allowing a correct evaluation of the content taken in by the reader; among other purposes. The analysis, based on functionalist presuppositions, rediscusses the conceptual limits between modality and evidentiality. It investigates, in terms of both quality and quantity, the use of evidentiality marks in academic dissertations. This genre was chosen due to the supposition that it is one of the textual genres whose source of information is mostly not the author himself. This fact influences the use of evidentiality marks in relation to the degree of the producerâs (un)commitment to the information provided. The 290 ocurrencies were collected from a corpus which comprises 10 academic dissertations in Humanity Studies. Only the introduction and the conclusion sections of these Works were considered. The results showed that the sense effects are associated to the use of evidentiality expressions in the kind of argumentation process in dissertations. This proves that the evidentiality in the academic discourse, particularly in the corpus collected, is used especially to assure a lower degree of commitment, being the verb the linguistic means most chosen to express this, and the inferential and quotative evidentiality notions as the most frequent ones.

This dissertation presents the findings of an on-line survey completed by 222 FYC (First Year Composition) instructors at universities and community colleges across the United States along with supplemental information derived from multiple open-ended interviews with seven FYC instructors in Arizona. Both survey and interview questions were designed to accomplish three primary goals: to determine which conventions of academic English FYC instructors identify as most important; to understand the common problems encountered by instructors in teaching those conventions, and; to solicit instructors' perceptions about ways in which learning outcomes might be improved. Results indicate general consensus among FYC instructors on which skills are both the most critical to academic English proficiency and the most difficult for their students to learn. At the same time, the survey and interview responses reflect widespread dissatisfaction with the ways in which academic English sequences are currently structured, apparently related to the instructors' common perception that the sequences are only "somewhat" successful in terms of preparing students for successful academic writing. Accordingly, the overwhelming majority of FYC instructors suggest changes for increasing the effectiveness of their programs; however, there is surprisingly little agreement among them on what those changes should be. The concluding section of this study presents pragmatic suggestions - congruent with a number of the instructors' observations - for reconfiguring FYC sequences. Additionally, it is argued that, aside from the targeted skills addressed by the instructors, the survey and interview responses indicate that academic English has been implicitly invested with culture-specific values which should be made explicit in instruction and which, given the gatekeeping status of FYC courses, the increasing diversity of student populations and the growing divide between the academic and wider cultures, require critical examination.

No description available.

This dissertation analysed the interactions between instructors and students in a language classroom in Russia. Using video-recorded data, instructor interviews, and student assessments from English classes at a private language school for children in the Urals region of Russia, a Conversation Analytic [CA] framework was employed to determine: how the talk (specifically turn-taking, adjacency pairs, and repairs) was sequentially organised; if and how the institutional setting constrained the talk; and if previously determined 'universal' structures of talk applied to this Russian academic discourse. This research also tested the hypotheses that the 'universal' structures of talk would apply regardless of gender, but would be used differently by the boys and girls, and by the instructors interacting with them. The relevance of the participants' institutional identities or gender to the interaction was also examined. The analysis showed that the participants did orient to their institutional identities of instructor or student, and the institutional setting did constrain the organisation of talk. The instructors' responses to the interviews and 'student assessment' questionnaires showed that they generally had positive attitudes toward girls and mixed attitudes toward boys. While the underlying sequences, the universal 'rules' of interaction, applied to interactions with both boys and girls, how (and how frequently) the sequences were used did vary by gender (i.e., typically 'male' and 'female' speech styles). Also, some of the organisation of talk showed that the instructors did orient to the students' genders in the classroom. This research is significant as the first CA study of the sequential organisation of talk in an institutional setting in Russia. In general, this research contributes to the CA findings on the organisation of talk in different languages, cultures, and settings; specifically, it provides the first point of comparison of Russian classroom interactions, from a CA perspective, with the large corpus of data already collected in classrooms in the Western tradition of education. Finally, this research is significant as it provides a thorough microanalysis of the relativity of gender-specific verbal behaviour; the analysis also shows how the instructors behave verbally, and in this way produce gender-specific communication styles. / Slavic Linguistics

Esta tese focaliza o ensino de língua espanhola numa região do Centro-Sul do estado do Paraná, Brasil. Trata-se de uma pesquisa de caráter discursivo e etnográfico, que se estrutura a partir de uma dupla articulação, a que se denomina políticas e práticas linguísticas. No que tange à primeira articulação, analisam-se textos produzidos na universidade e textos regulamentadores do ensino, notadamente, parâmetros, orientações e diretrizes curriculares, para investigar como as políticas contidas nestes documentos contemplam a realidade multilíngue do Brasil. Para tanto, apoia-se nas perspectivas discursivas propostas por Bourdieu (1998) e por Authier-Revuz (1998; 1999). No que concerne à segunda articulação, detém-se no estudo de práticas linguísticas escolares, com o intuito de compreender como ocorre a gestão das línguas em um contexto em que convergem o ucraniano e o português como línguas veiculares, e a língua espanhola como a língua estrangeira que se ensina no currículo escolar. Em vista disso, elabora-se uma discussão sobre o conceito de alternância linguística em contextos de aprendizagem, a partir das perspectivas defendidas por Nussbaum (1998) e Py (2004), bem como sobre a gestão das línguas como reflexo do desenvolvimento de uma competência plurilíngue (MOORE, 2009). Os resultados desta pesquisam apontam que, embora as produções e publicações atuais evidenciem um movimento em direção ao estudo das realidades multilíngues, a política linguística que atravessa estes textos ainda pauta-se em ideais monolíngues. Em relação às práticas linguísticas escolares, observou-se que os alunos manifestam um saber elaborado das línguas, ao alternarem entre o português, o ucraniano e o espanhol, e ao negociarem seus conhecimentos linguísticos com as docentes. No entanto, não reconhecem as suas práticas como plurilíngues. Assim, pois, defende-se a necessidade de uma reconceitualização do ensino de línguas, tanto ao problematizar os discursos hegemônicos sobre as línguas, como ao trabalhar para que práticas linguísticas plurilíngues sejam legitimadas, sobretudo, na escola. / This thesis focuses on the Spanish language teaching in the south-central area of the Paranas state, Brazil. Its a discursive and ethnographic research which is structured from a double articulation called politics and linguistics practices. Regarding to the first articulation, texts produced in the universities are analyzed, as the curricular guidelines and orientations, in order to investigate how the politics contained in these documents contemplates the Brazilian multilingual reality. The analysis is supported by the discursive perspectives proposed by Bourdieu (1998) and by Authier-Revuz (1998; 1999). Regarding to the second articulation, it is focused on the study of school linguistic practices, in order to understand how the languages management happens in a context which is used Portuguese and Ukrainian are used as vehicular languages and the Spanish language as a foreign language included in the school curriculum. As a result of this, it is prepared a discussion about the concept of linguistic switches, from the perspectives defended by Nussbaum (1998) and Py (2004), and about the languages management as a reflex of the multilingual competence development (MOORE, 2009). The results of this research point to, though the current productions and publications that show a movement in to the multilingual realities study, the linguistic politic contained in these texts, still, maintain themselves in the monolingual ideal. Concerning the school linguistic practices, it was observed that the students show an elaborated knowledge about the languages, when they switch among Portuguese, Ukrainian and Spanish and, also when they negotiate their linguistic knowledge with the teachers. But they dont recognize their practices as multilingual. Thus, there is a need of language teaching reconceptualization when discussing the hegemonic discourses about the languages, as well when working in order to legitimize the multilingual linguistic practices, mainly, at school.

Conselho Nacional de Desenvolvimento CientÃfico e TecnolÃgico / A evidencialidade à uma categoria linguÃstica que diz respeito à fonte das informaÃÃes veiculadas e, nessa indicaÃÃo, està tambÃm relacionada com estratÃgias de (des) comprometimento do enunciador com tais informaÃÃes. No intuito de investigar a manifestaÃÃo dessa categoria na construÃÃo dos trabalhos acadÃmicos de grau do portuguÃs brasileiro contemporÃneo, a pesquisa objetiva: a) rediscutir o estatuto da evidencialidade como categoria linguÃstica; b) analisar as relaÃÃes entre modalidade e evidencialidade; c) verificar as estratÃgias utilizadas no uso das marcas de evidencialidade nas monografias, dissertaÃÃes e teses; d) identificar, descrever e analisar as marcas linguÃsticas evidenciais nos trabalhos cientÃficos em foco; e) relacionar o uso de evidenciais com estratÃgias para efeito de (des)comprometimento na construÃÃo textual. Ao identificar e interpretar as marcas evidenciais na construÃÃo dos textos acadÃmicos, a pesquisa busca contribuir para a explicitaÃÃo dos efeitos de sentido vinculados à veiculaÃÃo das informaÃÃes de forma estratÃgica, jà que essas marcas sÃo utilizadas com propÃsitos diversificados, como, por exemplo, o recurso ao chamado âargumento de autoridadeâ, a atenuaÃÃo da responsabilidade em relaÃÃo ao que à dito, a modalizaÃÃo no contÃnuo entre a certeza e a nÃo-certeza, sugerindo um grau de comprometimento em relaÃÃo à verdade da proposiÃÃo, bem como um posicionamento crÃtico em relaÃÃo a ela. A anÃlise, orientada por pressupostos funcionalistas, conta com uma dimensÃo teÃrica, voltada para a rediscussÃo dos limites conceituais entre as categorias modalidade e evidencialidade; e uma dimensÃo analÃtica, que, em constante diÃlogo com a teoria, investiga, qualitativa e quantitativamente, o uso de marcas da evidencialidade nas monografias, dissertaÃÃes e teses. A escolha desses gÃneros textuais para a constituiÃÃo do corpus justifica-se pela suposiÃÃo de que apresentam uma maior quantidade de informaÃÃo cuja fonte nÃo à o prÃprio autor, o que condiciona o uso das marcas de evidencialidade na relaÃÃo observÃvel com o grau de comprometimento do produtor relativamente à informaÃÃo veiculada. A obtenÃÃo das 1500 ocorrÃncias para esta pesquisa concretizou-se a partir da organizaÃÃo de um corpus constituÃdo por trinta trabalhos acadÃmicos de grau (dez monografias, dez dissertaÃÃes e dez teses), dos quais foi utilizado um recorte como material de anÃlise, selecionando-se a introduÃÃo, um capÃtulo de fundamentaÃÃo teÃrica e a conclusÃo. Os dados obtidos foram analisados segundo as seguintes categorias: a) os meios linguÃsticos de expressÃo da evidencialidade no discurso acadÃmico que envolve a produÃÃo textual da monografia, da dissertaÃÃo e da tese foram especificados em relaÃÃo ao tipo de marca evidencial (item lexical ou gramatical - verbo, substantivo, adjetivo, advÃrbio, preposiÃÃo; enunciado metadiscursivo, justaposiÃÃo simples, normas citativas da ABNT; e à posiÃÃo de tal marca em relaÃÃo à fonte (anteposta, intercalada, posposta); b) os tipos de noÃÃes evidenciais caracterizadoras do aspecto semÃntico, a fonte da informaÃÃo (sujeito-enunciador, fonte externa definida, fonte externa indefinida, domÃnio comum); o acesso evidencial (direto, menos direto, indireto); a natureza evidencial da informaÃÃo (subjetiva, experiencial, inferencial, relatada); c) os aspectos pragmÃtico-discursivos, ou seja, os efeitos de sentido na construÃÃo da argumentaÃÃo relacionados à expressÃo do alto, mÃdio ou baixo comprometimento com o que està sendo informado. Predominou, na construÃÃo dos trÃs gÃneros de trabalhos acadÃmicos de grau, o efeito de baixo comprometimento, sendo a diferenÃa entre alto e baixo comprometimento aproximadamente a mesma nos trÃs gÃneros. Quanto aos aspectos semÃnticos, predominou a fonte externa definida, indireta e relatada, manifestada por meio de verbos dicendi, que se intercalam entre a fonte e o conteÃdo asseverado. / Evidentiality is a linguistic category related to the source of informations and also to strategies of commitment degrees of the author with these informations. With the intent of investigating the manifestation of that category in the construction of academic degree works in contemporary Brazilian Portuguese, this research aims to: a) re-discuss the status of evidentiality as a linguistic category; b) analyze the relations between modality and evidentiality; c) examine the strategies employed in the use of evidentiality marks in the monographs, dissertations and theses; d) identify, describe and analyze the evidential linguistic marks in the scientific works in focus; e) relate the use of evidentials to strategies for compromise effect in textual construction. By identifying and interpreting the evidential marks in the construction of academic texts, this research seeks to contribute to express explicitly the meaning effects linked to the transmitting of information strategically, since these marks are used for diverse purposes, such as resorting to the so-called "authority argumentâ, attenuation of liability in relation to what is said, modalization in the continuum between certainty and non-certainty, suggesting a degree of commitment in relation to the truth of the proposition, also a critical positioning in relation to the proposition. This analysis, guided by functionalist assumptions, has a theoretical dimension, aimed at re-discussion of the conceptual boundaries between modality and evidentiality categories, and an analytical dimension, which, in constant dialogue with the theory, researches in a qualitative and quantitative way the use of evidentiality marks in monographs, dissertations and theses. The choice of these text genres for the corpus constitution is justified by the assumption that they present a greater amount of information whose source is not the author himself, conditioning the use of evidentiality marks in the observable relationship with the degree of commitment of the author related to the information conveyed. The acquisition of 1500 instances for this research came true from the organization of a corpus consisting of thirty academic degree (ten monographs, ten dissertations and ten theses), which was used as a cut-off for analysis: introduction, one chapter of theoretical discussions and conclusion. Data were analyzed according to these categories: the type of mark evidential (lexical or grammar item â verb, noun, adjective, adverb, preposition; metadiscursive statement, simple juxtaposition, quotatives marks of ABNT) and the position of such mark in relation to source (before, between, after); b) the types of concepts characteristic of the evidential semantics are described, the source of information (the utterer, defined external source, undefined external source, common domain); the evidential access (direct, less direct, undirect); the nature of evidential information (subjective, experiential, inferential, reported); c) the pragmatical and discursive aspects, that is, the effects of meaning associated with the construction of the arguments in that concerns to the expression of high, medium or low commitment to what is being reported. In three genres of academic degree, the low commitment was more frequent and the differences between high and low commitment is nearly the same in these genres. Regarding to semantic aspects, the external, indirect and reported source was the most frequent, explicit by dicendi verbs, that appear between the source and the contend asserted.

4 Abstract The present thesis focused on the differences in the usage of hedging in research articles of native speakers of Czech writing in English and native speakers of English. Hedging is considered to be a fundamental part of any academic text ensuring that the results and outcomes of one's research are successfully communicated to its intended audience. The main aim was to discover, by analysing the most heavily hedged, concluding sections of research articles, how the native Czech speakers employ hedging in comparison to native English speaking authors and if they project features typical for Czech academic discourse into their work written in English. The most important outcome of the results of this study is the observation that one of the most salient features of Czech academic discourse, which is the low degree of authorial presence and interactivity of the text, does significantly influence the way the Czech authors present the outcomes of their research in English. This is evidenced by the significant difference in the frequency of the reader-oriented hedging strategies in the NCS corpus.

This thesis aims at providing a literature review of the academic debate regarding the use of childhood iconography in the humanitarian sector. This topic is not new and has been broadly discussed by academics since decades. Therefore, by using the alreadyexisting literature on the subject, this paper provides an in-depth understanding of the main arguments and controversies that have been made regarding the use of such images in the humanitarian field. In order to do so, it analyzes two famous pictures which produced considerable discussion in the research field: “The Struggling Girl” taken by Kevin Carter in South Sudan, in 1993 and “Alan Kurdi” photographed on a Turkish beach by Nilüfer Demir, in 2015. Overall, the study of the discourses generated by these two photographs exemplifies the diverging opinions of scholars on the topic and sheds light on the power relations and stigmas that persist within the humanitarian sector.

The subject of the present study is a comparison of two postmodifying constructions in noun phrases - finite relative adnominal clauses with the subject gap and nonfinite participial clauses - in spoken academic discourse: natural and social science. The comparison is based on the fact that both constructions realize the same clause element, i.e. a postmodifier in a noun phrase. The aim of the present study is thus to present major characteristics and functions of finite relative clauses and their reduced counterparts with respect to their distribution across the two subregisters of spoken academic discourse. The study is divided into three main parts: the theoretical background (Chapter 2) which defines the major characteristics and functions of the two postmodifying constructions and their mutual relationship, as are presented in the literature; the main part (Chapters 3 and 4) which provides the analysis of finite relative clauses and participial postmodifiers; and conclusion in Chapter 5.

[pt] Esta pesquisa tem por objetivo estudar a redação de vestibular enquanto gênero discursivo, visando caracterizar sua estrutura textual e enfocá-lo como um processo social dentro do contexto pedagógico. Buscando embasamento teórico na lingüística sistêmico-funcional (Halliday, 1994, Martin, 1989), nos estudos de gênero (Marchuschi, 2002) e em teorias de argumentação (Perelman, 1993, Breton, 1998), coletou-se um corpus de 135 redações, selecionadas de acordo com as notas a elas atribuídas. Os textos foram escritos por candidatos a cursos de diferentes áreas de três instituições de ensino superior, duas públicas e uma particular, localizadas em diferentes Estados do país. Como parte da análise foram observados os parâmetros de configuração textual do gênero através da identificação de tipos textuais e da descrição de estruturas argumentativas presentes nos textos. A redação de vestibular foi também estudada em termos de processos sociais, a partir da análise da sua relação com o contexto sócio-histórico e do uso de marcas de subjetividade e identidade discursivas presentes nos textos. A análise possibilitou ainda a caracterização do gênero estudado face ao discurso acadêmico, situando-o em uma esfera mais ampla, ou seja, partindo de sua materialização nos textos e relacionando-o ao contexto acadêmico. Os resultados da pesquisa mostram que os candidatos utilizam em suas redações diferentes tipos textuais, bem como diversos tipos de argumentos, sendo freqüentes as narrativas e os argumentos que indicam afetividade. Estas características, bem como o uso recorrente de um esquema simplificado de argumentação, o uso constante explicitação do contexto (Biber, 1988, Oliveira, 2002) e o uso pouco constante de nominalizações que possibilitam a expressão de idéias abstratas em detrimento de processos, expressos por verbos que retratam ações (Halliday,1994), parecem contribuir para a caracterização dos textos como pouco acadêmicos, frustrando as expectativas dos avaliadores, e levando-os a serem avaliados pelas bancas como pouco eficientes. Por outro lado, o uso de argumentos criativos, dotados de humor e considerados como eficientes, bem como o uso reduzido de marcas de subjetividade demonstram estarem os textos em consonância com exigências do discurso acadêmico. Esta pesquisa, portanto, mostra ser esse um gênero que apresenta ao mesmo tempo vários movimentos constitutivos inerentes ao discurso acadêmico e aspectos dissonantes, típicos de um texto não acadêmico, sendo por isso relevante o estudo de sua caracterização discursiva e papel social, com vistas ao ensino de escrita no contexto pedagógico brasileiro. / [en] The objective of this research is to study university composition entrance examinations as a discourse genre, in order to characterize its textual structure and to examine it as a social process in the pedagogic context. Theoretical aspects of systemic-functional linguistics (Halliday, 1994, Martin, 1989), genre studies (Marchuschi, 2002) and argumentation principles (Perelman, 1993, Breton, 1998) were applied to a corpus of 135 assignments. The texts, selected according to a range of grade variation, were written in Portuguese by applicants for courses of several areas from three universities located in different Brazilian states, two public institutions and a private one. As part of the analysis, textual configuration parameters were observed through the identification of the use of text types and argumentative structures found in the compositions. The analysis also aimed at the characterization of that particular genre in terms of academic discourse, leading it beyond its materialization in texts into the wider perspective of its relationship with the academic context. Research results show that applicants use different text types in their assignments, as well as several kinds of arguments, narratives being frequent as well as arguments with some emotional appeal. These characteristics, as well as the recurrent use of an oversimplified argumentative scheme and constant context explicitation (Biber, 1988, Oliveira, 2002) and with the constant use of nominalizations that refer to abstract ideas rather than processes, expressed through verbal forms that refer to actions (Halliday, 1994), point towards academic discourse. seem to contribute towards the characterization of these texts as not totally fitting academic discourse, thus frustrating examiners´ expectations and being evaluated by the committees as inadequate, due to lack of efficiency. On the other hand, the use of creative and humorous arguments, which are considered efficient, with reduced use of subjectivity marks also demonstrate that the texts analyzed are in accordance with the requirements of the academic context. Therefore, this research leads into the conclusion that university composition entrance examinations show both constitutive aspects inherent to academic discourse and also some dissonant characteristics typical of nonacademic texts. These results reinforce the need for the discursive characterization of this genre and its study as a social process in the Brazilian pedagogic context.