Lam Lai-chuen. / "May 2003." / Thesis (Ph.D.)--Chinese University of Hong Kong, 2003. / Includes bibliographical references (p. 261-280). / Electronic reproduction. Hong Kong : Chinese University of Hong Kong, [2012] System requirements: Adobe Acrobat Reader. Available via World Wide Web. / Electronic reproduction. Ann Arbor, MI : ProQuest Information and Learning Company, [200-] System requirements: Adobe Acrobat Reader. Available via World Wide Web. / Mode of access: World Wide Web. / Abstracts in English and Chinese.

Machine Translation (MT) for language pairs with long distance dependencies and word reordering, such as German-English, is prone to producing output that is lexically or syntactically incoherent. Statistical MT (SMT) models used explicit or latent syntax to improve reordering, however failed at capturing other long distance dependencies. This thesis explores how explicit sentence-level syntactic information can improve translation for such complex linguistic phenomena. In particular, we work at the level of the syntactic-semantic interface with representations conveying the predicate-argument structures. These are essential to preserving semantics in translation and SMT systems have long struggled to model them. String-to-tree SMT systems use explicit target syntax to handle long-distance reordering, but make strong independence assumptions which lead to inconsistent lexical choices. To address this, we propose a Selectional Preferences feature which models the semantic affinities between target predicates and their argument fillers using the target dependency relations available in the decoder. We found that our feature is not effective in a string-to-tree system for German-English and that often the conditioning context is wrong because of mistranslated verbs. To improve verb translation, we proposed a Neural Verb Lexicon Model (NVLM) incorporating sentence-level syntactic context from the source which carries relevant semantic information for verb disambiguation. When used as an extra feature for re-ranking the output of a German-English string-to-tree system, the NVLM improved verb translation precision by up to 2.7% and recall by up to 7.4%. While the NVLM improved some aspects of translation, other syntactic and lexical inconsistencies are not being addressed by a linear combination of independent models. In contrast to SMT, neural machine translation (NMT) avoids strong independence assumptions thus generating more fluent translations and capturing some long-distance dependencies. Still, incorporating additional linguistic information can improve translation quality. We proposed a method for tightly coupling target words and syntax in the NMT decoder. To represent syntax explicitly, we used CCG supertags, which encode subcategorization information, capturing long distance dependencies and attachments. Our method improved translation quality on several difficult linguistic constructs, including prepositional phrases which are the most frequent type of predicate arguments. These improvements over a strong baseline NMT system were consistent across two language pairs: 0.9 BLEU for German-English and 1.2 BLEU for Romanian-English.

The aim of this paper is to study whether there is a correlation between how subject pronouns are introduced in Spanish teaching and the pupils' difficulties regarding both their use and the understanding of verb conjugation. What is being investigated is: a) to what extent and in what contexts subject pronouns appear in Spanish textbooks, and b) how secondary education teachers of Spanish work with subject pronouns in the classroom. My conclusion is that the subject pronouns are under-represented in Spanish textbooks, probably due to the interpretation of the null subject parameter and Spanish as a pro-drop language. As far as the work with subject pronouns in the classroom is concerned, a more in-depth analysis needs to be carried out in order to be able to demonstrate a reliable correlation between the teaching method and the students' results. / Syftet med detta arbete är att studera om det finns ett samband mellan hur subjektspronomina introduceras i spanskaundervisningen och elevernas svårigheter kring både användandet av dessa och förståelsen för verbböjning. Det som undersöks är: a) i vilken utsträckning och i vilka sammanhang förekommer subjektspronomina i läromedel för undervisning i spanska, och b) hur högstadielärare i spanska arbetar med subjektspronomina i klassrummet. Min slutsats är att subjektspronomina finns underrepresenterade i läroböcker i spanska, förmodligen på grund av den tolkning som görs av nollsubjektsparametern och spanska som pro-drop språk. Vad gäller arbetet med subjektspronomina i klassrummet behöver en mer fördjupad analys genomföras för att det ska kunna påvisas ett tillförlitligt samband mellan undervisningsmetod och elevernas resultat.

O presente trabalho visa a contribuir com as pesquisas sobre histórias em quadrinhos, mais especificamente na área aplicada à articulação e codificação de sentidos presentes nesse gênero discursivo. Como arcabouço teórico pertinente à análise do discurso de base enunciativa, emprego nesta pesquisa os conceitos de semântica global, práticas intersemióticas e competências, presentes nas obras Gênese dos discursos (2008) e Análise de textos de comunicação (2011), de Maingueneau, além de retomar a reflexão de Foucault sobre as relações entre palavras e imagens, apresentada em Isto não é um cachimbo (2008). Do campo dos estudos aplicados às HQs, ressalto aqueles realizados por Will Eisner (2010) e Sonia Luyten (2000 e 2001), entre outros. A partir desses referenciais teóricos o presente estudo propõe investigar a materialidade do discurso verbo-imagético das histórias em quadrinhos, tendo como referencial verbal as onomatopeias e as mímesis contidas no corpus selecionado, o primeiro volume da primeira edição brasileira do mangá Cavaleiros do Zodíaco (2000), e compreender como a relação entre esses elementos e os planos discursivos do estrato imagético produz sentidos a serem decodificados pelos leitores. Por meio de tal análise, procuro evidenciar a necessidade de que seja considerada a existência e relevância de uma competência imagética no ato da leitura de histórias em quadrinhos, referindo-se especificamente à decodificação dos dispositivos de produção de sentidos sistematicamente empregados no estrato imagético desse gênero discursivo / This study aims at contributing to the researches about comics, more specifically those applied to the articulation and codification of meanings present in this genre. As relevant theoretic framework relevant to the enunciative basis discourse analysis, I apply in this research the concepts of global semantic, intersemiotic practices and competences, as shown in the books Gênese dos discursos (2008) and Análise de textos de comunicação (2011), written by Maingueneau, and retake the Foucault's reflection about the relationship between words and images, presented in Isto não é um cachimbo (2008). From the studies about comics I emphasize those performed by Will Eisner (2010) and Sonia Luyten (2000 and 2001), among other researchers mentioned throughout the dissertation. By these theoretic frameworks, this study proposes to investigate the materiality of the comics verb-imagetic discourse, considering as verbal referential the onomatopoeias and the mimesis contained in the selected corpus, the first issue of the first Brazilian edition of the manga Cavaleiros do Zodíaco (2000), and understand how the relationship between these elements and the discoursive plans from the imagetic stratum produces meanings to be decoded by readers. Through this analysis, I seek to highlight the need to consider the existence and the relevance of a imagetic competence in the act of reading comics, referring specifically to the decoding of the devices that produce meanings and are systematically employed in their imagetic stratum

Cette thèse a pour objectif d’étudier certains aspects du traitement et de l’acquisition de la morphologie verbale du russe. Le but de ce travail est double. Premièrement, nous avons étudié le traitement d’une alternance consonantique, la palatalisation, par des adultes russophones. Ce processus morphonologique mène à l’allomorphie des radicaux dans plusieurs classes verbales dont les verbes subissent la surgénéralisation dans le langage des enfants. Deuxièmement, nous avons testé l’effet de la rétroaction négative présentée dans l’input, surtout l’effet durable, sur l’élimination des erreurs de surgénéralisation fréquentes chez les enfants en russe. Dans la première étude, nous présentons des données expérimentales sur le traitement des emprunts et des non-mots. Plus particulièrement, cette étude vise à répondre à la question de savoir comment la palatalisation de consonnes dentales et vélaires est traitée par des locuteurs adultes du russe. Les résultats montrent que la palatalisation est semi-productive en fonction des facteurs suivants: a) la distribution des allomorphes à l’intérieur du paradigme, et b) la productivité des classes verbales. Nous supposons que la différence dans le traitement de la palatalisation chez les adultes devrait être reflétée dans le langage des enfants. Notre deuxième article présente les résultats de l’étude sur les effets de la rétroaction négative dans l’acquisition de la morphologie flexionnelle en russe. Pour ce faire nous avons mené une série de tâches induites auprès d’enfants russophones âgés de 3 à 4 ans. Des verbes sensibles à la surgénéralisation en yod /j/, une erreur typique des enfants de cet âge, ont été utilisés comme stimuli. Les participants ont été divisés en quatre groupes selon le type de rétroaction (correction, question de clarification et répétition) auquel ils étaient exposés. Dans chaque groupe de participants, nous avons observé une amélioration significative en production cible de formes verbales avec le temps. Cependant aucune différence significative n’a été trouvée concernant le type de rétroaction. Ces résultats suggèrent que la rétroaction négative ne joue pas un rôle important dans le processus d’acquisition. Ensemble, les deux études représentent une nouvelle contribution à la discussion sur les processus irréguliers en morphologie et le phénomène de surgénéralisation, ainsi que sur le (non) rôle de la rétroaction dans l’élimination des formes surgénéralisées dans le langage des enfants. / This thesis aims at studying certain aspects of Russian verb morphology processing and acquisition. The goal was two-fold: first, we investigated the productivity of morphonological alternations that lead to irregular verb stem allomorphy among adult speakers of Russian. The verbs in the study are known to undergo overregularization in Russian child speech. Second, we tested the (potentially) lasting effect of negative feedback on the retreat from overregularization errors in children. In the first paper, we present experimental data on the processing of loanwords and nonce words that focus on a morphonological alternation (palatalization) in Russian. This study addresses the issue of how stem allomorphy involving palatalization of the velar/palatal and dental/palatal types in the Russian verb system is processed by adults. Processing of palatalization is shown to be quite variable and to depend on: (i) different distribution of allomorphs (past/non-past or 1Sg./other forms) within the verb paradigm, and (ii) overall productivity of verbal classes. We also hypothesized that these differences should be reflected in child language verb morphology acquisition. The study presented in the second article investigates negative feedback effects on inflectional morphology acquisition in Russian. With that goal in mind, we conducted a series of elicited tasks with Russian speaking children aged from 3 to 4 years. Verbs which undergo overregularization in the non-past tense resulting from applying the yod /j/-pattern (typical errors for children of this age) were used as stimuli. Four groups of participants were formed accordingly to three types of feedback (Correction, Clarification question and Repetition), and a control group without feedback. Our results revealed a significant effect of time on target verb form production. However, no significant difference was observed as a function of feedback type, or even where there was no feedback. This finding supports the general hypothesis that negative feedback is not an important factor of language acquisition. Altogether, the results presented in this thesis provide new insights on irregular processes in Russian verb morphology, as well as on the inefficiency of negative feedback in the acquisition of L1 morphology.

nÃo hà / Esta pesquisa està centrada no estudo de produÃÃes textuais de alunos de 6 ano do ensino fundamental II, cujo gÃnero pertence ao conto popular. A abordagem dada ao tema teve como base pressupostos da LinguÃstica Textual, com Ãnfase no protÃtipo narrativo de Adam (1992; 2008) e no ISD de Bronckart (2007), especificamente, no tipo de discurso narraÃÃo. O objetivo à analisar a infraestrutura desses textos, verificando a presenÃa da sequÃncia narrativa ou do script, alÃm da construÃÃo do texto com as categorias: tempos verbais, organizadores temporais e pronomes. A pesquisa apresenta um carÃter teÃrico-prÃtico e à delineada em quase-experimental com um corpus que se compÃe de quarenta e duas (42) produÃÃes textuais divididas em produÃÃo inicial (PI) e produÃÃo final (PF). Essas produÃÃes textuais foram recolhidas durante um perÃodo de trÃs semanas com atividades, que tratavam das categorias citadas anteriormente, em forma de oficinas à luz das sequÃncias didÃticas de Schneuwly; Dolz (2004). A anÃlise levou-nos a observar que muitas das produÃÃes foram estruturadas segundo a sequÃncia narrativa de Adam (1992; 2008), outras, porÃm, apresentaram somente o script, conforme Bronckart (2007), comprometendo a infraestrutura textual de gÃneros do narrar como o que nos propomos a construir com os alunos do 6 ano. Apreendemos que os alunos possuem a estrutura narrativa internalizada, porÃm nÃo sabem distinguir entre o processo de intriga e o script, produzindo textos que apenas enumeram aÃÃes. Ao construÃ-los utilizam-se das categorias elencadas, priorizando os tempos verbais do narrar, apesar de em algumas situaÃÃes essas construÃÃes parecerem equivocadas. Os organizadores temporais, segunda categoria analisada, funcionam nesses textos como um elemento que auxilia os tempos verbais com as locuÃÃes adverbiais que se apresentam com maior incidÃncia. Na categoria pronomes, as construÃÃes apresentam anÃforas que demonstram o uso de um mecanismo da lÃngua, evitando repetiÃÃo de elementos no texto, o que constitui para um aluno em processo de desenvolvimento da escrita uma estruturaÃÃo de ideias. Acreditamos que a pesquisa constitui-se em uma reflexÃo teÃrica sobre a construÃÃo do texto narrativo de alunos de 6 ano. Esperamos ter contribuÃdo para compreensÃo da importÃncia da diferenciaÃÃo entre um texto narrativo e um script. A relevÃncia dessa pesquisa està na compreensÃo dessas categorias na infraestrutura do texto que auxiliam o aluno a produzir gÃneros do narrar e ao professor compreender esse processo, valorizando e auxiliando o aluno na construÃÃo da produÃÃo textual narrativa. / This work aims at analyzing the text productions of the II level fundamental sixth grade, concerned to the genre of popular narratives. The approach given to the theme is based on Text Linguistics, emphasizing Adamâs (1992; 2008) narrative prototype and Bronckartâs (2007) sociodiscursive interactionism, particularly about the type of narrative discourse. The objective is to analyze the infrastructure of these texts, verifying the presence of the narrative sequence or of the script, as well as of the constructions with the following categories: verb tenses, temporal organizers and pronouns. This is a theoretical-practical research, which can be described as an almost experimental one, whose corpus is formed by forty-two (42) text productions, divided into initial production and final production. These text productions, which were done by workshops in the light of Schneuwly and Dolzâs (2004) didactic sequences, were collected in a period of three weeks, containing activities with the categories previously cited. We observed that a lot of productions were structured according to Adamâs (1992; 2008) narrative sequences. However, other texts presented only the script, according to Bronckart (2007), what can jeopardize the text infrastructure of narrative genres as we proposed to construct with these students of the sixth grade. We noticed that the students have the narrative structure internalized, but they can not distinguish the plot process from the script, producing texts which only enumerate actions. When they build them, they use the mentioned categories, prioritizing narrative verb tenses, although some of these situations appear to be wrong. The temporal organizers, which are the second category analyzed, have the function of an auxiliary of the verb tenses with the most frequent adverbial locutions. Referring to the pronouns, the constructions present anaphors which show the use of a language mechanism that avoids the repetition of text elements. This results in the constitution of a writing development process and an idea structure process by the students. We believe that this research consists in a theoretical reflection about the construction of narrative texts by students of the sixth grade. We hope to contribute to the importance of differing a narrative text from a script. The relevance of this research is defined by the comprehension of these categories to the text infrastructure, which help the students to product narrative genres, and the teachers to comprehend this process, valuing and helping the students in the narrative text production.

Nesta tese, que se propôs a explorar as gramáticas brasileiras do português publicadas no século XIX, tivemos três objetivos. O primeiro foi realizar um mapeamento exaustivo da produção gramatical brasileira do português do século XIX, o qual foi feito por meio de fontes secundárias e catálogos de bibliotecas. Tivemos conhecimento de 127 títulos e 77 reedições, o que resulta em duzentas gramáticas. A partir dos dados coletados, que, em geral, eram título, subtítulo, autor, ano de publicação, local e casa impressora, formalizamos unidades de análise para, ao relacioná-las, por meio da metodologia da História Serial (Furet 1991, Barros 2005), realizar um mapeamento descritivo e interpretativo dessa produção. Nosso segundo objetivo, tendo selecionado, com base em quatro critérios combinados (prestígio, emergência, institucionalização e diálogo), dezoito exemplares (16 títulos, 2 reedições) dos duzentos e quatro de que tivemos conhecimento, foi de realizar uma análise orientada para os conteúdos dessas obras, organizada em torno de uma possível rede conceitual. A hipótese dessa análise, baseada em Auroux (2009[1992]) e Swiggers (2010), é de que alguns conceitos poderiam ser eixo de uma rede conceitual devido à sua influência técnica e, eventualmente teórica, em uma determinada tradição de descrição linguística. Essa hipótese surgiu nos resultados de nossa pesquisa de mestrado, em que, ao notar a complexidade da dimensão técnica (cf. Swiggers 2004) de algumas obras dessa tradição, pensamos ser necessário procurar meios de tornar sua análise mais homogênea (cf. Polachini 2013, Polachini 2016). Para testá-la, examinamos o conceito-chave de \'verbo substantivo\' apresentado nas obras selecionadas, que são amiúde influenciadas pela tradição da grammaire générale francesa, ao menos até 1880, quando passa a haver também influência do método histórico-comparativo (cf. Maciel 1910, Nascentes 1939, Elia 1975, Cavaliere 2001, Parreira 2011, Polachini 2013). Ademais, foi necessário formalizar uma metodologia a fim de reconhecer a rede conceitual em torno do verbo substantivo. Primeiramente, definimos conceitos gramaticais como sendo formados por meio de processos classificatórios da cadeia falada. Tais processos têm como produto, por um lado, termos, e, por outro, conceitos. Estes últimos podem ser analisados intensionalmente, por meio de sua definição e seu lugar numa determinada taxonomia, ou extensionalmente, considerando o inventário de dados linguísticos eleitos como exemplos e eventuais ilustrações desses dados em uso. Cada conceito é, portanto, analisado por meio das seguintes categorias: definição, taxonomia, exemplos, ilustrações e por seus termos. Consideramos rede conceitual as relações travadas entre o conceito-chave, verbo substantivo, e conceitos que, na tradição da grammaire générale teriam já alguma relação com o verbo substantivo (cf. Chevalier 1968; Colombat 1992; Raby 2000; Bouard 2007), a saber: gramática e linguagem, metaclasses e classes de palavras, modificadores do nome, verbo, oração (seus elementos e sua organização) apresentadas nas obras. Essas relações podem ser de quatro tipos: (1) equivalência; (2) intersecção (3) subordinação; (4) base teórica comum. Finalmente, distinguimos no texto gramatical a descrição linguística, que chamamos de texto, e as ampliações, reflexões ou críticas realizadas sobre essa descrição, que chamamos de metatexto. Tal metodologia nos permitiu uma visão mais clara e homogênea da dimensão técnica das obras (e de seu eventual ecletismo), além de proporcionar uma visão mais complexa de continuidades e descontinuidades dessa produção, que contrasta com aquelas das revisões históricas. Ademais, os resultados apontam também para: relações entre agendas dos autores ao publicarem suas gramáticas e seu ecletismo. Nossa terceira proposta foi de que esta tese fosse não apenas uma historiografia, mas também uma epihistoriografia dessa tradição. Assim, apresentamos dois apêndices: um com informações catalográficas das 204 gramáticas de que tivemos conhecimento, organizadas cronologicamente; e outro com os paratextos das dezoito obras analisadas. / In this thesis, which approached the Brazilian grammars of Portuguese published in the 19th century, we had three objectives. The first one was to conduct a comprehensive mapping of the Brazilian grammatical production on Portuguese in the 19th century, which we carried out through secondary sources and library catalogs. We have acknowledged 127 titles and 77 re-editions, which total two hundred grammars. From the data collected, which generally consisted of title, subtitle, author, year published, publisher city and publisher name, we built standardized units of analysis so that, when analyzing their relations among one another, we conducted a descriptive and interpretative mapping of this production, adopting the methodology of Serial History (Furet 1991, Barros 2005). By selecting eighteen grammars (16 titles, 2 re-editions) out of the acknowledged two hundred works, based on four combined criteria (prestige, emergence, institutionalization and dialog), our second object was to carry out an analysis oriented to the content of these selected works, organized under a possible conceptual network. The hypothesis of this analysis, based on Auroux (2009[1992]) and Swiggers (2010), is that some concepts could work as an axis of a conceptual network due to their technical, and sometimes theoretical, influence on a specific tradition of linguistic description. This hypothesis was raised in the results of our masters research, in which, by realizing the complexity of some of this traditions grammars technical dimension (cf. Swiggers 2004), we thought it to be necessary to find ways of turning the analysis more homogeneous (cf. Polachini 2013, Polachini 2016). In order to test its effectiveness, we examined the key concept of verbo substantivo (literally: substantive verb), presented in the selected works, which are frequently influenced by the tradition of the French grammaire générale, at least until 1880, when the influence of the historical-comparative method started to take place (cf. Maciel 1910, Nascentes 1939, Elia 1975, Cavaliere 2001, Parreira 2011, Polachini 2013). Moreover, it was necessary to formalize a methodology in order to recognize the conceptual network around the verbo substantivo. First, we defined the grammatical concepts as being formed through classifying processes of spoken speech. These processes have as products, on one side, terms and, on the other, concepts. The latter can be analyzed intentionally, through its definition and its places in a specific taxonomy, or extensionally, considering the inventory of linguistic data considered as examples and any illustrations of these data in use. Each concept is, therefore, analyzed through the following categories: definition, taxonomy, examples, illustrations and by its terms. We considered as conceptual network the relations between the key concept, verbo substantivo, and concepts that, in the grammaire génerales tradition would already have some relation to the verbo substantivo (cf. Chevalier 1968; Colombat 1992; Raby 2000; Bouard 2007), such as: grammar and language, metaclasses and parts of speech, noun modifiers, verb, clause (its elements and organization) presented in the grammars. These relations can be of three types: (1) identity; (2) intersection; (3) subordination; (4) shared theoretical background. Finally, in the grammatical text, we distinguished the linguistic description, which we called text, from its expansions, reflections and critical comments on this description, which we called metatext. Such methodology allowed a clearer and more homogeneous view of the technical dimension of the works analyzed (and their often eclecticism), in addition to providing a more complex view of continuities and discontinuities of this production, which contrasts with those of the historical reviews. Furthermore, the results also point to: relations between the authors schedules when publishing their grammars and their eclecticism; and the metatexts role between teaching and linguistic reflection. Our third objective was to make this thesis not only a historiography but also a epihistoriography of this tradition. Thus, we presented two appendices: one with cataloging information of the two hundred grammars which we acknowledged, organized chronologically; and another one with the paratexts of the eighteen works analyzed.

Submitted by Rosivalda Pereira (mrs.pereira@ufma.br) on 2017-05-17T17:43:00Z No. of bitstreams: 1 ManuelaViana.pdf: 2426723 bytes, checksum: 8d061fd93ee18a67fdaaf8669c1cdf9a (MD5) / Made available in DSpace on 2017-05-17T17:43:00Z (GMT). No. of bitstreams: 1 ManuelaViana.pdf: 2426723 bytes, checksum: 8d061fd93ee18a67fdaaf8669c1cdf9a (MD5) Previous issue date: 2017-04-24 / The paper shows the analysis writing result in Portuguese as Second Language, performed by the deaf in Brazilian sign language as First Language in networks. The aim was to analyse the verb usage in Portuguese phrasal writing structure as the deaf Second Language in networks. Take as a guideline: How do the Deaf who have Brazilian sign language as First Language employ the verb in Portuguese writing in networks? To answer this question, we go first to theorists whose work deal with Brazilian sign language, Portuguese, Portuguese writing, Portuguese as Second Language, verb, networks, writings and deafness, highlighting authors as Almeida (2007), Bagno (2011), Felipe (2013), Peixoto (2004), Quadros (2004, 2006), Recuero (2011), Zeni (2010), among others who discuss about the theme. In the second place, we gathered together a group in the net WhatsApp of about 20 deaf users of Brazilian sign language as First Language. That group took part in ―daily chats‖ about ordinary subjects in the network WhatsApp. Those ―chats‖ written in Portuguese were picked up from the above mentioned network and established the research corpus that was analysed under the Phenomenology standard experienced in two moments: graphic idea analysis and nomothetics analysis (abstract and unrelated ideas. From those processes arose five different rates of analysis: Verb in the Infinitive, Auxiliar Followed by main Verb, Verb with Oral sign, Unfit Verb and Unfit Usage and Inflection Verb. The analysed results showed: the network WhatsApp is considered a tool to help the acquisition of Portuguese as Second language, the deaf acquire Portuguese language as Second Language when they use the web, having in their writing a suitable verb inflection to understand the sentence and in their writings in spoken Portuguese as oral signifying. Starting from the understanding\interpretation that the inflection and finding attitude brought us, we suggest new studies in order to make closer the relation between. Brazilian sign language as First Language and Portuguese as Second Language so that the deaf can establish an interactive linking more proficient in the society that has Portuguese as First Language. / O trabalho apresenta a análise da produção escrita em Língua Portuguesa como L2 desenvolvida por surdos usuários da Língua Brasileira de Sinais- LIBRAS como L1, nas redes sociais. O objetivo foi analisar o emprego do verbo na estrutura frasal do português escrito como segunda língua dos surdos usuários das redes sociais. Partimos da seguinte questão norteadora: De que modo os surdos que têm a LIBRAS como sua primeira língua empregam o verbo no português escrito nas redes sociais? Para responder à questão norteadora, recorremos, primeiramente, aos teóricos cujos trabalhos versavam sobre LIBRAS, língua portuguesa, português escrito, português como segunda língua, verbo, redes sociais, escrita e surdez, destacando-se autores como: Almeida (2007), Bagno (2011); Felipe (2013), Peixoto (2004), Quadros (2004, 2006), Recuero (2011), Zeni (2010), dentre outros teóricos que discutem o tema em destaque. Num segundo momento, realizamos a criação de um grupo na rede social WhatsApp composto por 20 (vinte) surdos usuários de LIBRAS como L1. Os componentes do grupo participavam de ―conversas diárias‖ sobre temas do cotidiano na rede social WhatsApp. Essas ―conversas‖, escritas em português, foram capturadas da referida rede social e constituíram o corpus da pesquisa, o qual foi analisado à luz da Fenomenologia, vivenciando dois momentos: o da análise ideográfica e o da análise nomotética. Nesse processo, emergiram das ―conversas‖ dos componentes do grupo 5 (cinco) categorias de análise: verbo no infinitivo, verbo auxiliar acompanhado de principal, verbo com marca de oralidade, verbo com emprego inadequado e verbo com flexão adequada. Os dados analisados revelam os seguintes resultados: a rede social WhatsApp é uma ferramenta a ser considerada para auxiliar na aquisição do português como L2; os surdos se apropriam do português como L2, quando usam a rede social da web, adquirindo na escrita uma flexão verbal adequada para a compreensão da sentença e se valendo, em seus escritos, até de registros próprios do português falado, como as marcas de oralidade. A partir da compreensão/interpretação que a atitude de reflexão e de desvelamento nos proporcionou, sugerimos novos estudos, no sentido de estreitar os laços entre a LIBRAS como L1 e o português como L2, para que o surdo possa estabelecer uma relação interativa mais proficiente na sociedade que tem o português como L1.

Om något är sant eller inte är upp till forskningen och vetenskapen att undersöka. Trots det kan mycket information upplevas som sanning även om det bara är en åsikt, egen erfarenhet eller myt. Den här studien undersöker hur privatpersoner online skapar sanningar utifrån en osanning. Undersökningen är baserad på en tråd från diskussionsforumet ”Flashback”. Tråden har en frispråkig och destruktiv diskurs om personer från Somalia. Analysen är baserad på olika sätt att få något att upplevas som trovärdigt, exempelvis modalitet, hänvisningar och objektivitet. Resultatet visar att de flesta åsikterna och upplevelserna är modifierade på olika sätt för att framstå som mer trovärdiga och om något är trovärdigt ökar det chansen att någon annan anammar samma åsikt eller åtminstone tror att den är sann. / Whether something is true or not is a matter for science and research to investigate. Although, a lot of information can pass by as true even though it might just be an opinion, personal experience or myth. This study investigate how private persons online create a truth out of something untrue. The investigation is based on a thread from the forum for discussion named ”Flashback”. The thread have an outspoken and destructive discourse about people from Somalia. The analysis is based on different ways to make something appear as reliable, for example modality, referencing and objectivity. The result shows that most opinions and experiences are in different ways modified to seem more reliable, and if something is reliabile it increases the chanses of someone else to acquire the same opinion, or at least believe it to be true.

[pt] O estudo do sujeito é de extrema importância na prática pedagógica do ensino da Língua Portuguesa. Muitas questões têm sido levantadas sobre a impropriedade com que as definições se aplicam aos exemplos e exercícios das gramáticas escolares. Nessa dissertação apresentamos uma revisão dessas definições, estabelecendo comparações com os exemplos dados nas gramáticas tradicionais, materiais didáticos e frases que surgem no uso não-padrão da língua. Nosso arcabouço teórico é pautado na abordagem funcionalista, na qual priorizamos a gramática de valências que coloca o sujeito gramatical como um complemento e a gramática sistêmica de M. A K. Halliday que apresenta o sujeito desempenhando papéis definidos pelos verbos. A partir das questões levantadas e da análise das definições dadas temos a convicção de que aquilo que a prática tradicional convencionou como modelo para o ensino de análise sintática precisa ser reavaliado, uma vez que verificamos severas contradições entre a teoria e prática. / [en] The study of the subject is of distinct importance for the pedagogical practice of the Portuguese language teaching. Many issues have been raised about the inadequacy of the definitions applied to the examples and exercises in grammars. In this dissertation, we present a review of these definitions, stablishing comparisons among the examples given in traditional grammars, teaching materials and sentences that appear in the non-pattern use of the language. Our theoretical framework is based on the functionalist approach, in which we have chosen the valency grammar which classifies the grammatical subject as a complement and the systemic grammar of M.A. K. Halliday which presents the subject performing papers defined by the verbs. Based on the issues raised and based on the analysis of the definitions we are convinced that that model which the traditional practice has established as the model for the teaching of syntatic analysis must be re-evaluated, since we have noticed many serious contradictions between the theory and the practice.