published_or_final_version / abstract / Education / Doctoral / Doctor of Philosophy

This study explores various aspects of the spelling and punctuation development of six Tohono O'odham (Papago) children during their third and fourth grade years. The data used, which was collected in a prior study, consisted of 215 texts (13,793 words) written in natural classroom settings as part of the teacher's ongoing writing curriculum. Field notes decribing what the subjects did as they wrote, as well as written interviews, supplemented the texts themselves. A number of specific features were examined to explore the subjects' use of various level of linguistic information about spelling. Those features included: initial letters, vowel phonemes, consonant digraphs, the letters C and G, consonant gemination, bound morphemes, and homophones. Spellings involving permutation (changes in letter order) and the letters E and Y as final markers, as well as those spellings which were real words, were also examined. Finally, differences between spellers, the subjects' use of punctuation and capitalization, and orthography in the classroom context were analyzed. A number of conclusions were drawn from the analysis. The subjects spelled most words conventionally. The more frequently a word appeared in the subjects' writing, the more likely it was to be spelled conventionally. Selected orthographic features varied widely in how conventionally they were spelled, with those which were less predictable or more abstract tending to be more difficult. Almost every spelling feature examined showed growth from third to fourth grade. The invented spellings of particular features tended to reflect understandable, logical processes. Invented spellings which were either real words or permutations of the intended word were common. There were differences between children not only in how conventionally they spelled but in the types of invented spellings they produced. Punctuation was more difficult than spelling for the subjects, and its use varied greatly between subjects. Capitalization was comparable to spelling in how conventionally it was used. Children used a variety of linguistic information and spelling strategies as they wrote. There was evidence that punctuation usage was driven (at least in some cases) by conscious hypotheses about how it works. Children's metalinguistic knowledge about orthography may or may not parallel their use of it.

Five experiments explored how we learn to read and recognise words with typical and atypical spelling-sound mappings and to generalize to novel words. In Experiment 1, adults learned to read pseudowords with typical or atypical pronunciations. There was some evidence that prior exposure to word meanings enhanced orthographic learning. However, interpretation was clouded by stimulus control problems that plague research using natural alphabets. In Experiment 2, an artificial orthography paradigm was developed to overcome these problems. Adults learned to read novel words written in novel symbols. Post-training, they could generalize, indicating extraction of individual symbol sounds. The frequency and predictability of symbol-sound mappings influenced learning and generalization, mirroring natural language findings. Experiment 3 found extended training to improve item recognition and generalization. In Experiment 4, pre-exposure to item sounds plus an object referent vs. item sounds provided equivalent benefit for orthographic learning. By the end of training, this was limited to items with low frequency unpredictable symbol-sound mappings. In Experiment 5, pre-exposure to novel definitions enhanced orthographic learning more than pre-exposure to item sounds, but by the end of training, both conditions were again equally beneficial.

This study was designed to determine the relative effectiveness of a group-study approach to the teaching of spelling as compared with an individual-study approach to the teaching of spelling in the seventh grade of a school with a large number of students of Latin American extraction.

Nesta dissertação tomamos por objeto de estudo a romanização, em português, de palavras árabes (e·g, termos, topônimos e antropônimos). Estudamos por que alguns nomes são grafados de maneiras diferentes em um ou outro jornal, por exemplo, analisamos esquemas existentes de romanização e propomos um sistema que seja não apenas conciso e regular, mas também abrangente e minucioso, podendo ser usado tanto por leigos quanto por acadêmicos. Os seis capítulos encontram-se divididos em duas partes. Na primeira parte, abordamos a relação entre transposição ortográfica (i·e, transcrição, transliteração romanização e latinização) e tradução; apresentamos a língua árabe e sua escrita; ressaltamos elementos pertinentes da ortografia do português brasileiro. Na segunda parte, cotejamos oito esquemas de romanização árabeportuguês, apontando suas (des)semelhanças; apresentamos nosso sistema de transposição ortográfica, em seus quatro módulos (leigo simplificado, leigo comum, acadêmico amplo e acadêmico estrito), incluindo exemplos de aplicação em nomes, expressões e um breve texto; concluímos antecipando os desdobramentos da proposta. / This master\'s thesis presents our research concerning the romanization of Arabic words (e·g, technical terms, toponyms and anthroponyms) into Portuguese. We have first studied why the way some nouns and proper names are written varies from source to source, then analysed some of the existing romanization schemes and, finally, proposed a system both concise and regular as well as comprehensive and detailed, one which is suitable for laymen and scholars alike. there are six chapters, distributed in two parts. In the first part, we see four orthographic transposition methods (transcription, transliteration, romanization and latinization) and how they relate to translation; the Arabic language and its writing script; and relevant aspects of the Brazilian Portuguese orthography. In the second part, we compare eight ArabicPortuguese romanization schemes, highlighting their (dis)similarities; introduce all four modules (laymans simplified, laymans common, scholars broad and scholars narrow) of our orthographic transposition system, including samples comprising names, expressions and a small text; and finish off with an assessment on how said proposal may unfold.

Esta tese tem por objetivos apresentar a edição criteriosa de manuscritos brasileiros seiscentistas; mostrar tendências gramaticais do português do século XVII, com base nesses manuscritos e em impressos coetâneos; analisar a retenção gráfica desses textos, muito mais próxima da escrita fonética do que da etimológica e, ainda, analisar o uso freqüente das conjunções coordenativas. A preocupação com o estado da língua portuguesa no século XVII justifica-se pela pouca quantidade de estudos voltados especificamente a esse período, através de fontes primárias, e por ser este um período em que a escrita, baseada de modo geral nos textos literários, é considerada pelos lingüistas de modo geral como pseudo-etimológica, contrariando fatos apresentados nas cartas e textos notariais seiscentistas. Os passos seguidos para a análise partiram da edição criteriosa de manuscritos seiscentistas brasileiros, da comparação com impressos e com gramáticas e tratados de ortografia de mesma época e do confronto com o que dizem os atuais especialistas em lingüística diacrônica. Os resultados obtidos permitemnos afirmar que a ortografia portuguesa do século XVII e o uso dos nexos de coordenação apresentam fortes características do português medieval, contrariando o que se pregava na época, em função da revolução filosófica desencadeada pelo Renascimento. Tal conclusão, a nosso ver, deve-se, especialmente, ao fato de a educação em Portugal, e conseqüentemente no Brasil, estar nas mãos da Companhia de Jesus, defensora da filosofia escolástica e opositora, portanto, ao pensamento racional e cartesiano dos países reformados, o que manteve Portugal afastado das grandes correntes filosóficas do século em questão. / The purpose of this thesis is to present a discerning edition of seventeenth Brazilian manuscripts; to demonstrate grammatical trends of Portuguese from the Seventeenth Century, based on these manuscripts and on coetaneous printed versions; to analyze the graphical retention of these texts, much more near to the phonetic writing than to the etymological one, and, also, to analyze the frequent use of the coordinative conjunctions in those texts. The concern with the Portuguese language condition in the Seventeenth Century is justified by the little quantity of studies specifically focused on that period, through primary sources, and because that is a period in which the writing, usually based upon literary texts, is considered by almost all linguists as pseudo-etymological, contesting facts presented in letters and in seventeenth century notarial texts.The steps followed for the analysis arose from the discerning edition of seventeenth Brazilian manuscripts, the comparison with printed versions and with grammar books and treaties on orthography from the same period and from the parallel with what the specialists on diachronic linguistic say. The obtained results allow us to affirm that the Portuguese orthography from the Seventeenth Century and the use of the coordination nexus present strong characteristics of medieval Portuguese, opposing to what was preached in that period, on account of the philosophical revolution broken out by the Renascence. Such conclusion, in our point of view, is specially due to the fact that the Education in Portugal, and consequently in Brazil, in the hands of \"Companhia de Jesus\", defensor of the scholastic philosophy and, therefore, opponent of the Rational and Cartesian thought of the reformed countries, what kept Portugal away from the great philosophical chains of the century under discussion.

The diploma thesis deals with the orthography problems and the use of Czech language overview worksheets for the 5th grade of primary school. The thesis is divided into three parts. The theoretical part contains the definition of the subject matter in RVP ZV, the definition of orthography, its meaning, spelling functions and principles, types of spelling exercises, their classification and correction. A discussion of new changes and approaches in contemporary orthography is also in the theoretical part. The research part of the thesis contains a pilot verification of the Czech language overview worksheets and asks several questions. The aim was to find out the pupils' attitude to orthography, to reveal their most common spelling mistakes and to verify whether there was an improvement in the spelling while pupils were using the Czech language overview worksheets for the 5th grade of primary school. The practical part of the thesis contains created overview sheets for the 5th grade of elementary school.

Esta tese trata da transposição ortográfica de elementos morfossintáticos do árabe padrão moderno (APM) para o português brasileiro (PB) e constitui uma continua-ção de nossa dissertação de mestrado (Sano, 2011). Nesse trabalho anterior, trata-mos da transposição ortográfica de consoantes no mesmo sentido (APM PB) e desenvolvemos um sistema de romanização com vistas à transcrição e à translitera-ção do APM, ou seja, uma romanização que permitisse recuperar tanto a fala (os fo-nemas) quanto a escrita (os grafemas) da língua de partida. Nesta etapa atual, utili-zamos esse sistema de romanização para analisarmos dois corpora (a gramática de Cowan e o resumo da referida dissertação), com o intuito de identificar os pontos a serem abordados por um sistema de romanização que trabalhe além do nível da pa-lavra. Após esse levantamento, atualizamos o sistema de romanização anterior-mente desenvolvido, agora com o escopo estendido para o nível frasal, e reescreve-mos ambos os corpora aplicando as modificações propostas. / This dissertation deals with the orthographic transposition of morphosintactical elements from Modern Standard Arabic (MSA) into Brazilian Portuguese (BP), being a continuation of our masters thesis (Sano, 2011). In that previous work, we dealt with the orthographic transposition of consonants in the same pair of languages (MSA BP) and developed a romanization system meant to allow for the transcri-ption and transliteration of MSA, i.e., a romanization suitable for recovering both the original speech (the phonemes) and the original writing (the graphemes) of the star-ting language. In the current stage, we have used that system to analyse two corpora (Cowans grammar and our thesis abstract), in order to identify topics to be addres-sed by a romanization system inteded to be used beyond word level. After that, the now extended romanization system was used to rewrite both corpora, applying the proposed modifications encompassing the sentence level.

Dialectal Arabic speech research suffers from the lack of labelled resources and standardised orthography. There are three main challenges in dialectal Arabic speech recognition: (i) finding labelled dialectal Arabic speech data, (ii) training robust dialectal speech recognition models from limited labelled data and (iii) evaluating speech recognition for dialects with no orthographic rules. This thesis is concerned with the following three contributions: Arabic Dialect Identification: We are mainly dealing with Arabic speech without prior knowledge of the spoken dialect. Arabic dialects could be sufficiently diverse to the extent that one can argue that they are different languages rather than dialects of the same language. We have two contributions: First, we use crowdsourcing to annotate a multi-dialectal speech corpus collected from Al Jazeera TV channel. We obtained utterance level dialect labels for 57 hours of high-quality consisting of four major varieties of dialectal Arabic (DA), comprised of Egyptian, Levantine, Gulf or Arabic peninsula, North African or Moroccan from almost 1,000 hours. Second, we build an Arabic dialect identification (ADI) system. We explored two main groups of features, namely acoustic features and linguistic features. For the linguistic features, we look at a wide range of features, addressing words, characters and phonemes. With respect to acoustic features, we look at raw features such as mel-frequency cepstral coefficients combined with shifted delta cepstra (MFCC-SDC), bottleneck features and the i-vector as a latent variable. We studied both generative and discriminative classifiers, in addition to deep learning approaches, namely deep neural network (DNN) and convolutional neural network (CNN). In our work, we propose Arabic as a five class dialect challenge comprising of the previously mentioned four dialects as well as modern standard Arabic. Arabic Speech Recognition: We introduce our effort in building Arabic automatic speech recognition (ASR) and we create an open research community to advance it. This section has two main goals: First, creating a framework for Arabic ASR that is publicly available for research. We address our effort in building two multi-genre broadcast (MGB) challenges. MGB-2 focuses on broadcast news using more than 1,200 hours of speech and 130M words of text collected from the broadcast domain. MGB-3, however, focuses on dialectal multi-genre data with limited non-orthographic speech collected from YouTube, with special attention paid to transfer learning. Second, building a robust Arabic ASR system and reporting a competitive word error rate (WER) to use it as a potential benchmark to advance the state of the art in Arabic ASR. Our overall system is a combination of five acoustic models (AM): unidirectional long short term memory (LSTM), bidirectional LSTM (BLSTM), time delay neural network (TDNN), TDNN layers along with LSTM layers (TDNN-LSTM) and finally TDNN layers followed by BLSTM layers (TDNN-BLSTM). The AM is trained using purely sequence trained neural networks lattice-free maximum mutual information (LFMMI). The generated lattices are rescored using a four-gram language model (LM) and a recurrent neural network with maximum entropy (RNNME) LM. Our official WER is 13%, which has the lowest WER reported on this task. Evaluation: The third part of the thesis addresses our effort in evaluating dialectal speech with no orthographic rules. Our methods learn from multiple transcribers and align the speech hypothesis to overcome the non-orthographic aspects. Our multi-reference WER (MR-WER) approach is similar to the BLEU score used in machine translation (MT). We have also automated this process by learning different spelling variants from Twitter data. We mine automatically from a huge collection of tweets in an unsupervised fashion to build more than 11M n-to-m lexical pairs, and we propose a new evaluation metric: dialectal WER (WERd). Finally, we tried to estimate the word error rate (e-WER) with no reference transcription using decoding and language features. We show that our word error rate estimation is robust for many scenarios with and without the decoding features.

Made available in DSpace on 2016-04-28T20:57:09Z (GMT). No. of bitstreams: 1 Maura Spada Zanella.pdf: 2152042 bytes, checksum: 9aa09ea46d387f1b0ee4b182f0e58f1f (MD5) Previous issue date: 2007-04-16 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / Orthographic issues have long been bothering teachers and students, not only at initial levels of school years but also at more advanced levels and a reason for criticism and discrimination inside and outside school. In the learning process of written language, once the initial phase of mastering the alphabetic code is overcome, it is important that children realize that the right spelling of words depends not only on the relation of grapheme/phoneme, but also on orthographic conventions. In the pedagogical practice the existence of a relation between reading activities and the ability to spell words correctly is often admitted. It is accepted that manipulation of text through reading has a positive influence on the learning process of orthographic rules. Reviewing literature in the area we could not find empirical evidence to support this hypothesis. Results of studies show that skilled readers use different routes to identify words and may use two types of strategies: the direct form of lexical recognition or the indirect route which is based upon phonological decodification. The use of these two routes does not seem to be excludent, that is, they are both a common trait in a skilled reader and complement each other in an interactive process. However, there is evidence that readers can differ according to the predominant use of one of these strategies. This study verified in what way the associations between reading and orthographic ability, if they exist, would be also linked to the kind of reading strategy used, a more phonological or more semantic/lexical strategy. In this study the denomination semantic/lexical reader was adopted to refer to the reader who depends mostly on the global recognition of the word without applying a punctual decodification, and phonological reader to the one who analyses more precisely the phonology of the word decodifying it. This study involved 267 fourth-grade to sixth-grade students from a school in the city of São Paulo. Tasks were used to evaluate reading and orthography skills. Performances on such assignments were subject to statistical tests. The findings show significant differences between the writing performance of the reader who depends mostly on the global recognition of the word (semantic/lexical reader) and the reader who depends mostly on the phonology (phonological reader), the latter showing better performance. We can conclude that the reader s profile is an important factor when it comes to improving orthography. The phonological reader perceives the correct orthography of words when compared with the semantic/lexical readers, which favors the writing correct. We consider that this study has a pedagogical relevance once the results achieved may be used in reflections on orthography teaching in the context of Brazilian schools / As questões de ordem ortográfica há muito vêm incomodando professores e alunos, não apenas nos níveis iniciais da escolarização, mas se prolongando nos níveis mais avançados, sendo motivo de censura e discriminação na escola e fora dela. No processo de aprendizagem da linguagem escrita, superada a fase inicial do domínio do código alfabético, é importante que a criança perceba que a grafia correta das palavras depende não somente da relação grafema/fonema, mas de convenções ortográficas. Na prática pedagógica é freqüentemente aceita a existência de uma relação entre a atividade de leitura e a habilidade de ortografar corretamente as palavras. Admite-se que a manipulação de textos por meio da leitura tem influência positiva sobre a aprendizagem de regras ortográficas. Revisando a literatura da área não encontramos evidências empíricas que dêem suporte a essa hipótese. Resultados de pesquisas mostram que leitores hábeis fazem uso de diferentes rotas para identificar as palavras e podem utilizar dois tipos de estratégias: a forma direta de reconhecimento lexical ou a rota indireta que se apóia na decodificação fonológica. A utilização dessas rotas não parece ser excludente, ou seja, ambas estão presentes no leitor hábil e se complementam num processo interativo. Contudo há evidências de que os leitores podem diferenciar-se em função do uso predominante de uma dessas estratégias. A presente pesquisa verificou em que medida as associações entre leitura e habilidade ortográfica, se verificadas, estariam associadas também ao tipo de estratégia de leitura utilizada, se mais fonológica ou mais semântico-lexical. Na presente pesquisa adotou-se a denominação leitor semântico/lexical para o leitor que se apóia mais no reconhecimento global da palavra sem efetuar uma decodificação pontual e leitor fonológico o leitor que analisa mais detidamente a fonologia da palavra decodificando-a. A pesquisa foi realizada com 267 alunos de 4ª, 5ª e 6ª series do ensino fundamental de uma escola da rede municipal de ensino da cidade de São Paulo. Foram utilizadas tarefas para avaliar habilidades de leitura e ortografia. Os desempenhos nas tarefas foram submetidos a testes estatísticos. Os achados mostram diferenças significativas entre os desempenhos em escrita do leitor que se apóia mais no reconhecimento global da palavra (leitor semântico/lexical) e do leitor que se apóia mais na fonologia (leitor fonológico), tendo estes últimos melhores desempenhos que os primeiros. Conclui-se que o perfil do leitor é fator importante quando se trata de aprimorar a ortografia. O leitor fonológico percebe melhor a ortografia das palavras, quando comparado ao leitor semântico-lexical, o que favorece a escrita correta. Considera-se que o estudo possui uma relevância pedagógica na medida em que os resultados obtidos poderão vir a ser utilizados nas reflexões sobre o ensino da ortografia no contexto das escolas brasileiras