La thèse propose une étude des procédures de cohésion nominale et d’organisation discursive à l’écrit en français langue étrangère, à partir d’un corpus de textes narratifs et de type théorique écrits par vingt quatre étudiants de l’université de « Los Andes » au Venezuela dans différents moments de l’apprentissage de la langue. Une description des premières mentions des référents et des mécanismes anaphoriques [les densités et les modes de fonctionnement des formes] a été réalisée en y établissant les emplois spécifiques selon les types de discours repérés dans les textes. Cette analyse a été fondée d’une part sur les concepts de production langagière et de types discursifs de J.-P. Bronckart [1996], et, d’autre part, sur les travaux de M. Charolles [1991, 2002] sur l’anaphore et la référence, et, l’étude de G.de Weck [1991] sur la cohésion anaphorique. Par ailleurs, à partir d’un autre corpus constitué par des récits écrits par dix-neuf scripteurs français langue maternelle, une comparaison des performances cohésives des deux groupes des scripteurs a été effectuée. Celle-ci a permis de dégager les spécificités cohésives des textes des apprenants de français langue étrangère. Une analyse de la progression thématique mise en place selon la distribution des thèmes et des rhèmes [Combettes, 1988, 1993] des premières séquences des textes a été réalisée. Au terme de cette étude il a été possible d’établir les mécanismes de cohésion qui dépendent des conditions imposées par les types discursifs, par les consignes de la tâche d’écriture et - ou par les contenus thématiques des textes, et les stratégies utilisées par les apprenants en fonction de leur maîtrise de la langue. Finalement, des implications didactiques des résultats obtenus ont été mises en lumière. / This dissertation offers a study of the various nominal cohesion and discursive organisation procedures in written French as a Foreign Language based on a corpus of narrative and theoretical type texts written by twenty four students from the University of “Los Andes” in Venezuela at various stages of their language learning. A description of the first mentions of the referents and the anaphoric mechanisms [the densities and the modes of functioning of the forms] has been realized by making out the specific uses according to the type of speeches discovered in the texts. This analysis was based on the one hand on the concepts of linguistic production and discursive types of J.-P. Bronckart [1996], and, on the other hand, on M. Charolles’s researches [1991, 2002] on the anaphora and reference, and the study by G. de Weck [1991] on anaphoric cohesion. Moreover, based on another corpus constituted by narrative texts written by nineteen French native speakers, a comparison of the cohesive performance of both groups of writers has been made. It allowed us to draw out the cohesive specificities of the texts written by the French as a Foreign Language students. An analysis of the thematic progression set up according to the distribution of themes and rhemes [Combettes, 1988, 1993] of the first sequences of texts has been realized. Once this study was completed it was possible to establish the cohesion mechanisms which depend on conditions imposed by the discursive types, the instructions of the writing task and - or by the thematic contents of the texts, as well as the strategies used by the students according to their command of French. Finally, didactic implications of the results obtained have been highlighted.

Natural Language Processing (NLP) develops computational approaches to processing language data. Supervised machine learning has become the dominant methodology of modern NLP. The performance of a supervised NLP system crucially depends on the amount of data available for training. In the standard supervised framework, if a sequence of words was not encountered in the training set, the system can only guess at its label at test time. The cost of producing labeled training examples is a bottleneck for current NLP technology. On the other hand, a vast quantity of unlabeled data is freely available. This dissertation proposes effective, efficient, versatile methodologies for 1) extracting useful information from very large (potentially web-scale) volumes of unlabeled data and 2) combining such information with standard supervised machine learning for NLP. We demonstrate novel ways to exploit unlabeled data, we scale these approaches to make use of all the text on the web, and we show improvements on a variety of challenging NLP tasks. This combination of learning from both labeled and unlabeled data is often referred to as semi-supervised learning. Although lacking manually-provided labels, the statistics of unlabeled patterns can often distinguish the correct label for an ambiguous test instance. In the first part of this dissertation, we propose to use the counts of unlabeled patterns as features in supervised classifiers, with these classifiers trained on varying amounts of labeled data. We propose a general approach for integrating information from multiple, overlapping sequences of context for lexical disambiguation problems. We also show how standard machine learning algorithms can be modified to incorporate a particular kind of prior knowledge: knowledge of effective weightings for count-based features. We also evaluate performance within and across domains for two generation and two analysis tasks, assessing the impact of combining web-scale counts with conventional features. In the second part of this dissertation, rather than using the aggregate statistics as features, we propose to use them to generate labeled training examples. By automatically labeling a large number of examples, we can train powerful discriminative models, leveraging fine-grained features of input words.

This dissertation analyses seven modern Bible translations in French with respect to their renderings of Koine Greek participles. The sample consists of the Passion Story from the four Gospels (Matt 26-28, Mark 14-16, Luke 22-24 and John 18-21), and is comprised of 603 Hellenistic participles in all. The participle forms are studied in six categories according to their syntactic function. The comparison focuses on differences in translation strategy, i.e. formal equivalence, omission and different kinds of transpositions, with special attention given to the choice of verb form. There is a discussion of the adequacy of contemporary, explicative theories of systemic differences between the passé simple/passé composé and the imparfait. A large number of examples are analysed in detail. The results of the survey show that the most prominent differences in translation strategies concern the predicative participle. Furthermore, this was the category that occurred most frequently in the sample. The Catholic scientific and literary translation La Bible de Jérusalem (1998) is the most literal of the seven versions. A high level of formal equivalence is also registered in the other scientific translation, La Traduction Œcuménique de la Bible (1988), even though application of this strategy outweighs the use of finite verbs, that is to say, the most common transposition. La Bible en français courant (1996) is the least literal: generally, it transposes the participle of the source text with a finite verb. This transposition is also very frequent in the literary La Bible de la Pléiade (1971). Most of the omissions are recorded in the recent literary La Bible, Nouvelle traduction (2001), which is shown to be the most divergent translation. Omissions are also frequent in the pastoral La Bible des moines de Maredsous (1968) and the liturgical La Traduction liturgique de la Bible (1977). When translated in conjunction with an element comprising a verb in one of the non-indicative moods (infinitive, imperative, participle and subjunctive), both the present and the aorist predicative participles are, to a large extent, rendered by a simple form, expressing non-accomplishment. However, the Bible de Jérusalem stands out with its greater use of compound present participles than any other version. When the predicative participle of the source text is transposed with a verb in the indicative mood, the passé simple is generally used to render the aorist; for the present participle, the imparfait is more frequent than the passé simple. Nevertheless, here too the passé simple accounts for a significant portion of the equivalents, especially in the two translations where transpositions formed by finite verbs are particularly important. There exist a few cases where some translators chose to use the passé simple/passé compose, while others chose the imparfait. The various details, tables and linguistic analyses in this dissertation provide a solid basis for accurately characterizing the various modern attempts made at reproduce this ancient text – a text so often translated, paraphrased, interpreted and deeply integrated in our cultural heritage.

No description available.

Submitted by Edileide Reis (leyde-landy@hotmail.com) on 2013-05-13T15:27:51Z No. of bitstreams: 1 Maria Silva.pdf: 1245786 bytes, checksum: 47d4412aab645a8affa79e9db5506aad (MD5) / Approved for entry into archive by Alda Lima da Silva(sivalda@ufba.br) on 2013-06-04T17:28:02Z (GMT) No. of bitstreams: 1 Maria Silva.pdf: 1245786 bytes, checksum: 47d4412aab645a8affa79e9db5506aad (MD5) / Made available in DSpace on 2013-06-04T17:28:02Z (GMT). No. of bitstreams: 1 Maria Silva.pdf: 1245786 bytes, checksum: 47d4412aab645a8affa79e9db5506aad (MD5) Previous issue date: 2004 / Esta dissertação analisa, numa perspectiva sociolingüística, as estratégias de realização do objeto direto anafórico no dialeto rural afro-brasileiro, focalizando as variantes que distanciam o português do Brasil (PB) e o português europeu (PE): a categoria vazia e o uso do pronome lexical (ele/ela). Parte-se do pressuposto de que as diferenças existentes entre o PB e o PE se devem, não só a motivações internas à estrutura da língua, mas, principalmente, a motivações externas (sócio-históricas) decorrentes do contato entre línguas ocorrido, quando no Brasil conviveram os portugueses, os índios e os negros trazidos da África como escravos. Buscando avaliar o quanto o contato entre línguas afetou o desenvolvimento histórico da língua portuguesa no Brasil, o foco da pesquisa dirige-se para as comunidades rurais afrobrasileiras isoladas, do interior do Estado da Bahia. Por terem-se mantido até bem recentemente em relativo isolamento, tais comunidades devem guardar não só costumes culturais bastante antigos, mas também padrões lingüísticos, que permitam encontrar evidências de processos de variação e mudança resultantes do contato entre línguas. Dessa forma, o corpus analisado reuniu quatro comunidades localizadas em regiões diversas do Estado da Bahia, em que houve grande concentração de mão-de-obra escrava, a saber: Helvécia, no extremo-sul da Bahia; Rio de Contas, na Chapada Diamantina; Cinzento, no semi-árido; e Sapé, no Recôncavo Baiano. / Salvador

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / A presente investigação se volta para a expressão de objeto direto anafórico (ODA) na produção de dados espontâneos de aquisição bilíngue simultânea (BFLA Bilingual First Language Acquisition) de Português Brasileiro (PB) e de inglês. A literatura em BFLA tem assumido que as duas línguas são adquiridas de maneira independente (DE HOUWER, 1990, 2005; PARADIS; GENESEE, 1996), porém indica que pode haver momentos de interação entre os dois sistemas linguísticos, resultando em transferência entre propriedades paramétricas das línguas (HULK; MÜLLER, 2000; MÜLLER; HULK, 2001; PÉREZ-LEROUX, et al 2009 ; STRIK; PÉREZ-LEROUX, 2011; SORACE, 2011). Essa investigação é conduzida com base na teoria de Princípios & Parâmetros (CHOMSKY, 1981) do Gerativismo, reformulada com o Programa Minimalista (CHOMSKY, 1995). PB e inglês se diferenciam em relação à marcação paramétrica para ODA: o PB admite objeto nulo anafórico (ON) e o inglês não. A hipótese de trabalho adotada é de que haverá transferência do PB para o inglês. Dois tipos de categorias nulas são observados: objetos nulos dêitico (Odeit) e anafórico (ON). Assume-se que a manifestação do primeiro tipo indica um estágio default universal, que constituiria uma estratégia facilitadora (RIZZI, 2005). A possibilidade de uso de ONs agramaticais no inglês é concebida como resultado da presença de dados ambíguos, que reforçariam essa possibilidade equivocadamente no inglês, em consonância com o defendido em Hulk & Müller (2000) e Müller & Hulk (2001). Assume-se, ademais, que as restrições semânticas que regem a distribuição das formas possíveis para ODA no PB (CYRINO, 2006; LOPES, 2009) só poderiam ser detectadas em uma faixa etária mais alta. São analisados dados espontâneos de três bilíngues simultâneos (N, L e A) em interação com seus pais. N foi acompanhado dos 2;1,18 aos 3;8,24 anos de idade, enquanto L foi acompanhada dos 2;5,30 aos 3;1,1 anos e A foi acompanhado dos 3;2,6 aos 3;8,26 anos. As manifestações de ODA foram identificadas e classificadas em DP, pronome, ON e ODeit. Comparando os bilíngues, constatou-se que cada criança parece estar em um momento de aquisição: N apresenta instâncias de ODeit em contextos imperativos e pronomes aparecem apenas no inglês aos 2;5,2 anos. L tem preferência por DPs nas duas línguas e usa pronomes apenas na língua inglesa. Aos 2;6,22 anos, surgem instâncias de ON com mais frequência no PB, mas também no inglês. A criança A apresenta todos os tipos de preenchimento de ODA e ONs aparecem nas duas línguas. Os dados indicam que ONs agramaticais estão presentes no inglês, sugerindo que há transferência do PB para o inglês / This research focuses on the expression of the anaphoric direct object (ADO) in Bilingual First Language Acquisition (BFLA) of Brazilian Portuguese (BP) and English. It has been assumed in the BFLA literature that the two languages are acquired independently in BFLA context (DE HOUWER, 1990, 2005; PARADIS; GENESEE, 1996). However studies also indicate that there may be moments in the course of the bilingual acquisition in which both languages interact resulting in parameter transfering (HULK; MÜLLER, 2000; MÜLLER; HULK, 2001; PÉREZ-LEROUX, et al 2009 ; STRIK; PÉREZ-LEROUX, 2011; SORACE, 2011). The framework adopted is the Generative Theory, particularly, the Principle & Parameter Theory (CHOMSKY, 1981) reformulated within the Minimalist Program (CHOMSKY, 1995). BP and English set different parameters for ADO: BP allows for null objects whereas English does not. The hypothesis is that there will be transfer from BP to English. Two types of null categories are observed: deictic object (Odeit) and anaphoric object (NO). We assume that the expression of the first type of null category indicates a universal default stage which would configurate a facilitation strategy (RIZZI, 2005). The possibility of use of ungrammatical NOs in English is conceived as a result of ambiguous data, which wrongly reinforce the possibility of NOs in English, in accordance with Hulk & Müller (2000) and Müller & Hulk (2001). We claim that the semantic restrictions which govern the expression of ADO in BP (CYRINO, 2006; LOPES, 2009) could only be detected with older bilinguals. The spontaneous data of three simultaneous bilinguals (N,L and A) in interaction with their parents are analyzed. N was observed from 2;1,18 to 3;8,24 years old, L was observed from 2;5,30 to 3;1,1 years old and A,from 3;2,6 to 3;8,26 anos. The forms used for the expression of ADO were identified and classified into DP, pronoun, NO and ODeit. Comparing the three bilinguals, we verified that each child seems to be in a particular moment of acquisition: N presents instances of Odeit in imperative contexts and pronouns appear only in English at 2;5,2. L shows preference for DPs in both languages and pronouns pop up only in English. At 2;6,22 years old, instances of NO appear in BP, but also in English. The bilingual A presents all types of expression of ADO and NOs appear in both languages. The data indicates that ungrammatical NOs are present in English suggesting transfer from BP to English

Submitted by Viviane Lima da Cunha (viviane@biblioteca.ufpb.br) on 2016-09-23T12:22:42Z No. of bitstreams: 1 arquivototal.pdf: 3967529 bytes, checksum: 17e8000091e75546f18d4625ec646445 (MD5) / Made available in DSpace on 2016-09-23T12:22:42Z (GMT). No. of bitstreams: 1 arquivototal.pdf: 3967529 bytes, checksum: 17e8000091e75546f18d4625ec646445 (MD5) Previous issue date: 2009-12-22 / The co-reference is usually defined as a strategy of textual progression that recalls a previous entity, called antecedent, through the use of an anaphora. Different areas of science have studied the co-reference for its relevance to the local coherence, as well as to the discourse comprehension. However, a crucial question remains and requires to be clarified, that is about the cognitive mechanisms and the linguistics principles that underlie the choice of anaphora among the various forms of the language. The current work has its theoretical connection with Experimental Psycholinguistics which deals with the process of the phases, and more especifically with the processing of the co-reference. This work compares the efficiency of the lexical pronouns vs. repeated-name, as well as hiperonimos vs. hiponimos, both seen as ways of one's inter-sentential anaphoric recalling in Brazilian Portuguese. It was also verified that the wide range of the Centering Theory (Grosz, Joshi e Weinstein, 1983, 1995), which states a slower effect known as Repeated NP Penalty when recalling an antecedent using the repeated noun instead of a pronoun, as well as the The Informational Load Hypotheses (Almor 1990, 1999, 2000), as an alternative concept that connects the processing cost with the discourse function. For that, the current research used three on-line self paced reading experiments and analyzed its results by using both the T-Test and the ANOVA one. In the first experiment the lexical pronouns were read faster than the repeated pronouns, in accordance to the Centering Theory and the he Informational Load Hypotheses. The second experiment, the co-reference was better established by the hiperônimos instead of hipônimos, reinforcing the cost-funcion principle stated by Almor (1990, 1999, 2000), while the Repeated NP Penalty by Gordon et. al. (1993, 1995), is concerned about the dichotomy between lexical pronouns vs repeated nouns. The syntactic prominence was also discussed in Chamber's & Smith (1999) research as well as in Leitão's (2005), but the third experiment stated the independent act of the syntactic prominence in relation to the type of the anaphora, even though that does not exclude the influence of the structural parallelism. / A co-referência é cornumente definida como uma estratégia de progressão textual e caracterizada pela retomada de urna entidade prévia, também denominada antecedente, através de urna anáfora. Ela vem sendo estudada por várias áreas do conhecimento científico, devido a sua importância para a coerência local e, conseqüentemente, para a compreensão do discurso, mas uma questão crucial continua em aberto e requer maiores esclarecimentos: quais os mecanismos cognitivos e os princípios lingüísticos que subjazem a escolha da anáfora, entre a multiplicidade de formas existente na língua. O presente trabalho se insere no quadro teórico da Psicolingüística Experimental que trata do processamento de frases e, mais especificamente, do processamento da co-referência. Nele, compararnos a eficiência dos pronomes lexicais vs. nomes repetidos e dos hiperônimos vs. hipônimos, como formas de retomada anafórica inter-sentencial do sujeito em Português Brasileiro. Verificamos também a abrangência explicativa da Teoria da Centralização (Grosz, Joshi e Weinstein, 1983, 1995), que postula urn efeito de retardamento, mais conhecido corno Penalidade do Nome Repetido, ao retomar um antecedente proeminente sintaticamente usando urn nome repetido em vez de urn pronome, e da Hipótese da Carga Informacional (Almor 1990, 1999, 2000), com uma concepção alternativa que relaciona o custo de processarnento e atenção discursiva. Para isso, aplicamos três experimentos on-line de leitura auto-monitorada e validamos estatisticamente seus resultados através do Teste-T e da ANOVA. No primeiro experimento, os pronomes lexicais foram lidos mais rapidamente do que os names repetidos, em consonância corn a Teoria da Centralização e com a Hipótese da Carga Informacional. No segundo experimento, a co-referenda foi estabelecida rnais facilrnente pelos hiperônimos do que pelos hipônimos, ratificando o principio de otimização entre custo de processarnento e função discursiva, defendido por Almor (1990, 1999, 2000), enquanto a Penalidade do Nome-Repetido, constatada pioneiramente por Gordon et. al. (1993, 1995), limita-se a dicotornia pronornes lexicais vs. nomes repetidos. A proeminência sintática também foi questionada no estudo de Chambers e Smith (1999) em Inglês e de Leitão (2005) em Português Brasileiro, mas o terceiro e último experirnento comprovou sua atuação independente do tipo de anáfora, apesar de não descartar a influência do paralelismo estrutural.

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / A presente investigação se volta para a expressão de objeto direto anafórico (ODA) na produção de dados espontâneos de aquisição bilíngue simultânea (BFLA Bilingual First Language Acquisition) de Português Brasileiro (PB) e de inglês. A literatura em BFLA tem assumido que as duas línguas são adquiridas de maneira independente (DE HOUWER, 1990, 2005; PARADIS; GENESEE, 1996), porém indica que pode haver momentos de interação entre os dois sistemas linguísticos, resultando em transferência entre propriedades paramétricas das línguas (HULK; MÜLLER, 2000; MÜLLER; HULK, 2001; PÉREZ-LEROUX, et al 2009 ; STRIK; PÉREZ-LEROUX, 2011; SORACE, 2011). Essa investigação é conduzida com base na teoria de Princípios & Parâmetros (CHOMSKY, 1981) do Gerativismo, reformulada com o Programa Minimalista (CHOMSKY, 1995). PB e inglês se diferenciam em relação à marcação paramétrica para ODA: o PB admite objeto nulo anafórico (ON) e o inglês não. A hipótese de trabalho adotada é de que haverá transferência do PB para o inglês. Dois tipos de categorias nulas são observados: objetos nulos dêitico (Odeit) e anafórico (ON). Assume-se que a manifestação do primeiro tipo indica um estágio default universal, que constituiria uma estratégia facilitadora (RIZZI, 2005). A possibilidade de uso de ONs agramaticais no inglês é concebida como resultado da presença de dados ambíguos, que reforçariam essa possibilidade equivocadamente no inglês, em consonância com o defendido em Hulk & Müller (2000) e Müller & Hulk (2001). Assume-se, ademais, que as restrições semânticas que regem a distribuição das formas possíveis para ODA no PB (CYRINO, 2006; LOPES, 2009) só poderiam ser detectadas em uma faixa etária mais alta. São analisados dados espontâneos de três bilíngues simultâneos (N, L e A) em interação com seus pais. N foi acompanhado dos 2;1,18 aos 3;8,24 anos de idade, enquanto L foi acompanhada dos 2;5,30 aos 3;1,1 anos e A foi acompanhado dos 3;2,6 aos 3;8,26 anos. As manifestações de ODA foram identificadas e classificadas em DP, pronome, ON e ODeit. Comparando os bilíngues, constatou-se que cada criança parece estar em um momento de aquisição: N apresenta instâncias de ODeit em contextos imperativos e pronomes aparecem apenas no inglês aos 2;5,2 anos. L tem preferência por DPs nas duas línguas e usa pronomes apenas na língua inglesa. Aos 2;6,22 anos, surgem instâncias de ON com mais frequência no PB, mas também no inglês. A criança A apresenta todos os tipos de preenchimento de ODA e ONs aparecem nas duas línguas. Os dados indicam que ONs agramaticais estão presentes no inglês, sugerindo que há transferência do PB para o inglês / This research focuses on the expression of the anaphoric direct object (ADO) in Bilingual First Language Acquisition (BFLA) of Brazilian Portuguese (BP) and English. It has been assumed in the BFLA literature that the two languages are acquired independently in BFLA context (DE HOUWER, 1990, 2005; PARADIS; GENESEE, 1996). However studies also indicate that there may be moments in the course of the bilingual acquisition in which both languages interact resulting in parameter transfering (HULK; MÜLLER, 2000; MÜLLER; HULK, 2001; PÉREZ-LEROUX, et al 2009 ; STRIK; PÉREZ-LEROUX, 2011; SORACE, 2011). The framework adopted is the Generative Theory, particularly, the Principle & Parameter Theory (CHOMSKY, 1981) reformulated within the Minimalist Program (CHOMSKY, 1995). BP and English set different parameters for ADO: BP allows for null objects whereas English does not. The hypothesis is that there will be transfer from BP to English. Two types of null categories are observed: deictic object (Odeit) and anaphoric object (NO). We assume that the expression of the first type of null category indicates a universal default stage which would configurate a facilitation strategy (RIZZI, 2005). The possibility of use of ungrammatical NOs in English is conceived as a result of ambiguous data, which wrongly reinforce the possibility of NOs in English, in accordance with Hulk & Müller (2000) and Müller & Hulk (2001). We claim that the semantic restrictions which govern the expression of ADO in BP (CYRINO, 2006; LOPES, 2009) could only be detected with older bilinguals. The spontaneous data of three simultaneous bilinguals (N,L and A) in interaction with their parents are analyzed. N was observed from 2;1,18 to 3;8,24 years old, L was observed from 2;5,30 to 3;1,1 years old and A,from 3;2,6 to 3;8,26 anos. The forms used for the expression of ADO were identified and classified into DP, pronoun, NO and ODeit. Comparing the three bilinguals, we verified that each child seems to be in a particular moment of acquisition: N presents instances of Odeit in imperative contexts and pronouns appear only in English at 2;5,2. L shows preference for DPs in both languages and pronouns pop up only in English. At 2;6,22 years old, instances of NO appear in BP, but also in English. The bilingual A presents all types of expression of ADO and NOs appear in both languages. The data indicates that ungrammatical NOs are present in English suggesting transfer from BP to English

No description available.

Many L2 students in Namibia are not adequately prepared for the academic demands of university courses, mainly because of poor reading skills in the L2. University students reading below their maturational levels, can mainly be attributed to their print-impoverished backgrounds, as reading is a skill that develops mainly through reading. Without any assistance, poor readers at university will continue to read poorly and as a result perform weaker in their academic courses compared to their peers who are better readers. The overall aim of this study is to explore the effects of a reading intervention program on a group of university students in Namibia. A quasiexperimental method with a control and an intervention group was employed. The effect of reading ability on academic performance was also investigated. In addition, through a combination of quantitative and qualitative methods, the pre-literacy experiences of students, the differences between good and poor readers at university, their views about the reading intervention program as well as the attitudes and practices of university lecturers towards reading instruction at tertiary level were examined. / Linguistics / M.A. (with specialisation in Applied Linguistics)