The current study investigated the realization for the English voiceless postalveolar affricate /tʃ/ in Najdi Saudi ESL learners' production. The goal of the study was to investigate whether Najdi Saudi ESL learners have difficulties in pronouncing /tʃ/. Both linguistic and extra-linguistic factors were taken into consideration. Eighteen Najdi ESL learners' with different length of residency in the US were the sample of the study. Sixteen English words with /tʃ/ in initial and final position represented the instrument of the study. Data were analyzed using both SPSS and Speech analyzer software. The results showed that Najdi ESL learners have difficulties with /tʃ/, and they pronounced it as /ʃ/. Pronunciation for /tʃ/ was more difficult in word-final position than word-initial. The study also, found that learners with longer LORs produced more accurate pronunciation than learners with shorter LORs. Commonly used words were not a variable of interest, however the study showed that learners had fewer errors with commonly used words, than uncommon ones. Finally, the results provided support for theories and hypotheses such as the CAH (Lado, 1957), MDH (Eckman, 1977), and LTT (Gass and Selinker, 1994).

DOMINGOS, Francisca Liliane da Costa. Aprendizagem formal de fonologia e seus efeitos na pronúncia dos sons vocálicos do inglês de aprendizes brasileiros. 2012. 115f. – Dissertação (Mestrado) – Universidade Federal do Ceará, Departamento de Letras Vernáculas, Programa de Pós-graduação em Linguística, Fortaleza (CE), 2012. / Submitted by Márcia Araújo (marcia_m_bezerra@yahoo.com.br) on 2014-09-25T12:52:19Z No. of bitstreams: 1 2012_dis_flcdomingos.pdf: 1366301 bytes, checksum: 72dc8a4377ef83fd57c007e63ded24ac (MD5) / Approved for entry into archive by Márcia Araújo(marcia_m_bezerra@yahoo.com.br) on 2014-09-25T15:12:39Z (GMT) No. of bitstreams: 1 2012_dis_flcdomingos.pdf: 1366301 bytes, checksum: 72dc8a4377ef83fd57c007e63ded24ac (MD5) / Made available in DSpace on 2014-09-25T15:12:39Z (GMT). No. of bitstreams: 1 2012_dis_flcdomingos.pdf: 1366301 bytes, checksum: 72dc8a4377ef83fd57c007e63ded24ac (MD5) Previous issue date: 2012 / The purpose of this study was to investigate if, and how, the metalinguistic knowledge of the phonological system of the English language affects the level of accuracy of the production of English vowel sounds by prospective teachers of English and how this knowledge enables them to monitor their own pronunciation. Specifically, the study attempted to: identify the problems English learners have when pronouncing vowel phonemes in English; investigate the level of perception the prospective teachers of English have concerning their own difficulties and problems in producing vowel sounds; analyze if the prospective teachers of English begin to more precisely identify their own problems/difficulties in the production of the vowel phonemes once they gain metalinguistic knowledge about the phonological system of English while taking the course English Language Phonology; and, finally, analyze if, and how, the metalinguistic learning of the phonological system of English helps learners to monitor their pronunciation in both controlled activities in the course English Segmental Phonology (focussing on the production of sounds) and less controlled activities in courses which focus on communicating messages, especially after a time gap between the instructed learning and spontaneous communicative production. This study is theoretically supported by McLaughlin’s Model of Information Processing (1983), Fledge’s Model of Speech Acquisition (1995), Best’s Perceptual Assimilation Model (1995), Krashen’s Monitor Hypothesis (1972), and other works which attempt to apply models of learning and of language acquisition to distance learning. In order to achieve the above mentioned objectives, portfolio activities recorded by 29 (twenty nine) prospective English teachers of the Distance English Teaching Certificate Program, at UFC-UAB, were collected while the learners were taking the following courses: English Segmental Phonology, which focuses on the production of sounds of the language, and English V: Oral Comprehension and Production, which focuses on communicating messages. An analysis of the recordings indicated: i) the most recurrent problems in the production of oral sounds result from the improper substitution of English sounds for Portuguese sounds, either because students don’t perceive their distinction or because they are influenced by the graphonemic spelling of word; ii) the students were able to perceive and correct their own pronunciation problems, which suggests that instruction on the phonological system of English allowed the students to activate and use self-monitoring strategies; iii) as the students learned about the phonology of English, they began to identify and correct their own pronunciation problems; and iv) after some time of having taken the Phonology Course, students were able to identify and correct their pronunciation problems, maintaining a 70% level of accuracy in the production of the oral sounds, even in activities which focus was not on the production of the sounds but on message communication. These results suggest that metalinguistic instruction on the phonological system of English can be helpful in the development of self-monitoring strategies and, therefore, optimize the production of oral sounds in a foreign language. The long-lasting effect of metalinguistic knowledge of phonology in the pronunciation of English can also be implied. / Esse trabalho teve por objetivo, investigar se, e de que maneira, o conhecimento metalinguístico do sistema fonológico da LI influi no nível de acurácia da produção dos sons vocálicos de professores de inglês em formação e como este conhecimento lhes possibilita a melhor monitorar sua pronúncia. Mais especificamente, buscávamos: identificar os problemas de produção dos sons vocálicos do inglês apresentados por aprendizes dessa língua; investigar o nível de percepção dos professores de inglês em formação acerca de seus problemas na produção dos sons vocálicos da LI; analisar se, à medida que os professores em formação adquirem conhecimento metalinguístico sobre o sistema fonológico da LI durante a disciplina Fonologia da Língua Inglesa, passaram a identificar com mais precisão seus problemas/dificuldades de pronúncia dos fonemas vocálicos; e, finalmente, analisar se, e de que forma, a aprendizagem metalinguística do sistema fonológico da LI auxiliaria esses professores em formação a monitorar sua pronúncia, tanto em atividades controladas durante a disciplina de Fonologia Segmental da Língua Inglesa (foco na produção dos sons) como em atividades mais livres em disciplinas subsequentes (foco na produção de mensagem), principalmente após decorrido um certo espaço de tempo entre a instrução explícita e a atividade de produção espontânea. Como base teórica para tais objetivos, usamos o modelo de processamento da informação de Barry McLaughlin (1983), o modelo de aprendizagem da fala (FLEGE, 1995) e o modelo de assimilação perceptual (BEST, 1995), a teoria do monitor de Krashen (1972), além de tentarmos relacionar esses conceitos aplicados ao EAD. Para alcance dos objetivos mencionados, coletamos e analisamos atividades de portfólio gravadas por 29 (vinte e nove) professores em formação do curso semipresencial de Letras/Inglês da UFC/UAB, durante as disciplinas de Fonologia Segmental da LI, com foco na produção de sons, e Língua Inglesa V-A: Compreensão e Produção Oral, com foco na produção de mensagem. Nossos resultados foram os seguintes: os problemas mais recorrentes em relação à produção oral dos sons vocálicos pelos professores em formação se referem à substituição imprópria de sons, seja pela não percepção de certos sons da LI que não existem no PB, seja pela influência grafo-fonêmica das palavras na pronúncia dos aprendizes; quanto ao nível de percepção dos aprendizes em relação aos seus problemas de pronúncia, podemos dizer que muitos desvios de produção oral foram percebidos e corrigidos de maneira a sugerir que o automonitoramento foi ativado pela instrução explícita; à medida que adquiriram conhecimento explícito em Fonologia, os professores em formação foram capazes de identificar e corrigir seus problemas de pronúncia; e, após decorrido certo tempo da instrução explícita, os participantes da pesquisa mantiveram um nível de 70% (setenta por cento) em acurácia na produção dos sons vocálicos da LI, quando o foco da atividade era a produção de mensagem e não somente de sons. Tais resultados nos levam a atestar a afirmação de que a instrução explícita em Fonologia pode auxiliar no desenvolvimento do automonitoramento, e assim, na otimização da produção oral. Além disso, também podemos sugerir que o efeito da instrução explícita na produção oral é durável.

La littérature scientifique dédiée à l’accentuation en anglais australien standard contemporain (SAusE) est, contrairement à celle qui concerne la prononciation de ses voyelles, peu étendue. Après un chapitre introductif proposant le contexte historique dans lequel le SAusE est né et a été décrit, sa définition actuelle ainsi que sa description phonologique, notre étude est consacrée à un examen systémique de l’accentuation lexicale en SAusE. Un corpus test a été mis en place spécifiquement, qui comporte la quasi-totalité des verbes dissyllabiques, des préfixés pluricatégoriels et des exceptions aux règles d’accentuation des mots de deux syllabes et plus, ainsi qu’un large échantillon d’emprunts aux langues aborigènes. Ces quelques 3500 items ont été choisis précisément parce qu’ils font partie des mots les plus susceptibles de connaître une variation accentuelle en anglais contemporain. Notre approche se situe dans la lignée de Lionel Guierre et propose un traitement dictionnairique dans lequel les éléments du corpus sont tous étudiés au travers des données de chacune des éditions les plus récentes des Longman Pronouncing Dictionary, Cambridge English Pronunciation Dictionary et Macquarie Dictionary qui constituent les dictionnaires de références dans le domaine de la prononciation de l’anglais. Ces données ont été complétées, lorsque cela était nécessaire, par des données fréquentielles issues du Corpus of Contemporary American English et par des données orales enregistrées spécialement pour cette étude, selon une démarche similaire à la première partie du protocole du projet Phonologie de l’Anglais Contemporain. Notre analyse met en évidence une grande stabilité accentuelle intervariétale entre le SAusE, l’anglais britannique standard et l’anglais américain standard et propose un relevé étayé et détaillé des spécificités accentuelles lexicales du SAusE contenues dans le corpus étudié. / The scientific literature dedicated to word stress in contemporary standard Australian English (SAusE) is not very extensive contrary to the one dealing with the pronunciation of vowels in this variety of English. We will introduce the historical context in which SAusE first emerged and was described, its current definition as well as its phonological description. Our study is then devoted to the systemic study of lexical word stress in SAusE. A corpus was specifically put together for this particular research : it includes most of the dissyllabic verbs, of the prefixed multicategorial words and of the words that are exceptions to the rules of word-stress assignment in English, as well as a large sample of borrowings from Aboriginal languages. These 3500 items were not chosen randomly : they are words which specifically tend to show word-stress variation in contemporary English. Our approach follows Lionel Guierre’s and offers a dictionary treatment in which all the elements of our corpus are studied using the data of each of the most recent editions of the Longman Pronouncing Dictionary, the Cambridge English Pronunciation Dictionary and the Macquarie Dictionary, all reference books on English pronunciation. Frequency data extracted from the Corpus of Contemporary American English and oral data recorded for this study (following a protocol similar to that which is applied in the first part of the project Phonologie de l’Anglais Contemporain) were added to the dictionary data when necessary. Our analysis brings to the fore a great accentual intervarietal stability between SAusE, standard British English and standard American English, and offers a detailed and complete list of the lexical accentual specificities of SAusE ascertained through our study.

La présente thèse propose une description et une étude multidimensionnelles (sociolinguistiques, phonologiques et phonétiques) de la variété d'anglais parlée dans le Greater Manchester. Nous offrons une discussion sur les enjeux méthodologiques et épistémologiques de l'étude du changement linguistique et de l'utilisation des corpus en linguistique. Notre travail est mené dans le cadre du programme PAC (Phonologie de l'Anglais Contemporain : usages, variétés et structure) et au sein du projet LVTI (Langue, Ville, Travail, Identité) sur la base du corpus PAC-LVTI Manchester, constitué de données authentiques et récentes récoltées sur le terrain. Notre analyse se concentre notamment sur le phénomène de nivellement dialectal, qui a été l'objet de nombreuses recherches récentes en sociolinguistique anglaise. Nous nous intéressons en particulier à l'hypothèse de l'expansion d'une variété supralocale dans le nord de l'Angleterre. Notre étude concerne essentiellement les voyelles du Greater Manchester, et repose sur une analyse phonético-acoustique de la production des locuteurs de notre corpus. Nous relevons les caractéristiques majeures de la variété mancunienne, telles qu'elles ont pu être décrites dans les quelques travaux publiés jusqu'ici, et étudions leur corrélation avec des facteurs sociolinguistiques classiques comme l'âge, le genre, ou le niveau socio-économique. Nous explorons également la pertinence des facteurs attitudinaux pour l'étude de nos données. Sur la base de nos résultats phonético-acoustiques, nous nous prononçons en faveur de la pertinence de l'opposition de longueur en anglais, repensée en termes de poids vocalique au niveau de la représentation des voyelles, et proposons une modélisation du système phonologique de la variété du Greater Manchester dans le cadre de la phonologie de dépendance. Nous discutons des évolutions du système à la lumière du phénomène de nivellement dialectal, et nous interrogeons sur la pertinence des facteurs internes et externes pour les expliquer. / This thesis offers a multidimensional (sociolinguistic, phonetic, and phonological) description and study of the variety of English spoken in Greater Manchester. We discuss the study of linguistic change and the use of corpora in linguistics from a methodological and epistemological point of view. Our work is conducted in the framework of the PAC programme (Phonology of Contemporary English: usage, varieties and structure) and within the LVTI project (Language, Urban Life, Work, Identity), and based on the PAC-LVTI Manchester corpus, which is composed of authentic and recent fieldwork data. Our analysis notably focuses on the phenomenon of regional dialect levelling, which has been largely documented in recent English sociolinguistic research. In particular, we are interested in the hypothesis of the expansion of a supralocal variety in the north of England. Our study deals mainly with the vowels of Greater Manchester English, and relies on a phonetic-acoustic analysis of our informants' realisations. We describe the major characteristics of the Mancunian variety based on the few studies published so far, and statistically evaluate their correlation with traditional sociolinguistic factors such as age, gender or socio-economic profile. We also explore the relevance of attitudinal factors for the study of our data. On the basis of our phonetic-acoustic results, we speak in favour of the relevance of the opposition of length in English, which we reconsider in terms of vocalic weight in the representation of vowels. We then offer a model of the phonological system of Greater Manchester inspired by Dependency Phonology. We discuss the evolutions of the system in the light of regional dialect levelling, and question the role played by internal and external factors in these linguistic changes.

The topic of the diploma thesis is A Linguistic Analysis of a Baroque Sermon. The text, written by Václav Ignác Blovský, dates back to 1724. The diploma thesis continues in my bachelor work. It´s aim was to give the analysis of phonetic and morphological features of the same text. The diploma thesis is divided into two parts - the theory and the research. The first one deals with the author´s life and his work, the cult of St. John of Nepomuk and sermons dedicated to this Czech patron saint in general. The analytical part summaries the results of the previous research and includes the syntactic analysis. The aim of the diploma thesis is to find out the level of the Czech language in the early 18th century on the basis of the findings.

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / The present work has as main target to scrutinize errors of writing which motivations have phonetic/phonologic foundations. In order to do so, we take into consideration 35 texts, deriving of 27 pupils from two groups of the 8th grade from a state public school of the city of Santa Maria RS. For the data analysis we essentially apply the classification proposed by Zorzi (1998) which was adapted. In this manner, from the nine categories considered in this study, only one of them includes errors motivated by issues of the own orthographic system of Brazilian Portuguese, while the others contain phonetic/phonological aspects. Therefore, to each considered category, in a first moment, we try to explain in a phonological basis the data they contain. We also make comparisons with findings of other studies dealing with the relationship Phonology/Orthography. Inserted in a connectionist perspective, we present explanations for some of the occurrences found in the data. Finally, by the means of connectionist presuppositions, we advocate a possible change in the phonological system of the learner after the acquisition of writing. In this light, so to formalize such process, we present a Connectionist Optimality Theory analysis, focusing, especially, the acquisition of the lateral liquid in coda position. / O presente trabalho tem como principal objetivo analisar erros de escrita cuja motivação é de base fonética/fonológica. Para isso, partimos de 35 textos, oriundos de 27 alunos de duas turmas de 8ª série de uma escola pública estadual do município de Santa Maria-RS. Para a análise dos dados, utilizamos, essencialmente, a classificação proposta por Zorzi (1998), a qual foi adaptada. Desse modo, das nove categorias consideradas neste estudo, apenas uma delas abrange erros motivados por arbitrariedades do próprio sistema ortográfico do português brasileiro, enquanto as demais abarcam aspectos de cunho fonético/fonológico. Assim, para cada categoria considerada, buscamos, em primeiro momento, explicar os dados nela inseridos com base na fonologia da língua. Realizamos, também, comparações com achados de outros estudos que trataram da relação Fonologia/Ortografia. Inseridos em uma perspectiva teórica de cunho conexionista, apresentamos explicações para algumas ocorrências encontradas nos dados com base nos pressupostos de tal base teórica. Finalmente, por meio de pressupostos do Conexionismo, defendemos a possibilidade de mudança no sistema fonológico do aprendiz após a aquisição da escrita. Nesse sentido, a fim de formalizarmos tal processo, apresentamos uma análise a partir da Teoria da Otimidade Conexionista, enfocando, especialmente, a aquisição da líquida lateral na posição de coda silábica.

Made available in DSpace on 2014-06-11T19:26:48Z (GMT). No. of bitstreams: 0 Previous issue date: 2010-03-05Bitstream added on 2014-06-13T20:34:48Z : No. of bitstreams: 1 fonte_js_me_arafcl.pdf: 1050185 bytes, checksum: 62150efda4256f7a7a4f691b83b097d4 (MD5) / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) / Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP) / O objetivo deste trabalho é investigar as qualidades vocálicas do português arcaico (PA), período trovadoresco, a partir da consideração das rimas e da grafia das Cantigas de Santa Maria (CSM) de Afonso X, remanescentes da segunda metade do século XIII. Esta pesquisa partiu de hipóteses levantadas por estudos anteriores sobre as qualidades fonológicas das vogais tônicas, pretônicas e átonas finais do PA. A metodologia empregada no estudo das vogais tônicas e átonas finais do PA baseia-se, neste trabalho, no mapeamento de todas as ocorrências dessas vogais nas rimas das CSM. Ao observar as possibilidades e impossibilidades de rima, no corpus analisado, esta pesquisa trouxe informações sobre a realização fônica das vogais portuguesas, em um momento passado da língua, que não possui registros orais. Para o estudo das vogais pretônicas do PA, foram considerados os casos de variação gráfica entre <e> e <i>, e entre <o> e <u>, identificados nas CSM. Este trabalho mostrou que muitos dos contextos fonológicos responsáveis pelo levantamento de vogal pretônica, em algumas variedades do português brasileiro (PB) atual, também influenciavam o alçamento dessa vogal no PA. Além disso, esta pesquisa mostrou que é possível obter informações sobre a realização fonética das vogais pretônicas do português partindo de dados de escrita. A partir deste trabalho, foi possível obter um quadro dos sistemas vocálicos do PA, em posição tônica, pretônica e átona final. Esta pesquisa também trouxe algumas informações importantes sobre as transformações por que passaram as vogais do português ao longo da história da língua / This work aims to study the vowel qualities in Medieval Portuguese (MP), through the observation of rhymes and writing in Alfonso X’s Cantigas de Santa Maria, composed on the second half of the 13th century. First of all, this dissertation considered all information presented by previous researches concerning stressed, pretonic and unstressed word-final vowels, in MP. The study of stressed and unstressed word-final vowels, in this work, was based in the consideration of the rhymes in CSM. From the analyses of the possibilities and impossibilities of rhyme, in the corpus considered, it was possible to determine the phonetic realization of vowels in MP. Concerning pretonic vowel in MP, this dissertation analyzes the variation between <e> and <i>, and between <o> and <u>, in CSM. This research shows that many of the phonological contexts in which pretonic vowel raising occurs, in Brazilian Portuguese (BP), were already common in the pretonic raising in MP. Furthermore, it proves that it is possible to investigate the pretonic vowel phonetic realization, in Portuguese, through the consideration of writing data. Finally, this dissertation presents a phonological description of the stressed, pretonic and unstressed word-final vowel systems of MP. It also detects changes in vowel quality, from Latin to MP, and from this one to BP

Made available in DSpace on 2014-06-11T19:32:47Z (GMT). No. of bitstreams: 0 Previous issue date: 2008-09-17Bitstream added on 2014-06-13T20:24:05Z : No. of bitstreams: 1 borges_pr_dr_arafcl.pdf: 1045933 bytes, checksum: 73c85130f57b56862c59725d5c3b1c54 (MD5) / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) / Esta tese tem por objetivo o estudo da estrutura morfofonológica das formas verbais futuras do Português Arcaico (PA), no período conhecido por trovadoresco (fins do século XII até meados do século XIV), determinando se essas formas podem ser consideradas simples (uma única palavra fonológica) ou compostas ou perifrásticas (duas palavras fonológicas). Em outras palavras, objetiva-se determinar se, na época em questão, essas formas eram analíticas (formadas pelo infinitivo do verbo principal + verbo aver conjugado no presente do indicativo, no futuro do presente, ou pelo infinitivo do verbo principal + verbo aver/ir, no pretérito imperfeito do indicativo, no futuro do pretérito) ou sintéticas (cuja estrutura é radical + vogal temática + desinência modo-temporal + desinência número-pessoal). O corpus é composto das 420 Cantigas de Santa Maria de Afonso X, tomadas a partir da edição de Mettmann (1986a, 1988, 1989). A pesquisa parte do mapeamento e do levantamento quantitativo de todas as formas futuras presentes no corpus, classificando-as quanto ao tipo (aparentemente sintéticas ou indubitavelmente analíticas), relacionando os dados mapeados com a ocorrência de clíticos e o posicionamento do acento. Os resultados obtidos são analisados a partir de critérios para distinção de formas simples e compostas colhidos em estudos anteriores (MASSINI-CAGLIARI, 1999, 2006). Foi analisado, também, o grau de gramaticalização entre o verbo principal e o auxiliar de futuro, a partir dos critérios estabelecidos por Bybee, Pagliuca e Perkins (1991). Os resultados obtidos mostram que os futuros do presente e do pretérito em PA devem ser considerados formas analíticas, constituídas do infinitivo do verbo principal mais o verbo aver conjugado no presente do indicativo, para o futuro do presente, ou do infinitivo do verbo principal seguido do verbo... / The objective of this Dissertation is to study the morphophonological structure of the future verb forms in Archaic Portuguese (AP) spoken during the so-called archaic period (late 12th C until mid 14th C), in order to establish whether these forms can be considered simple forms (a single phonological word) or compound or periphrastic (two phonological words). In other words, our objective is to determine if at the studied period these forms were analytical (formed with the infinitive of the main verb + verb aver in the present, the future indicative, or by the infinitive of the main verb + aver/ir in the past imperfect) or synthetic (base form + theme vowel + mood-time inflection + person-number inflection). The corpus is formed of 420 Cantigas de Santa Maria by Afonso X edited by Mettmann (1986a, 1988, 1989). Firstly, all future forms found in the corpus were mapped, quantified and classified as to their type (apparently synthetic or positively analytic) and these data were related with the occurrence of clitics and the stress position. The results obtained were analyzed according to criteria used in former studies (MASSINI-CAGLIARI, 1999, 2006) to distinguish simple and compound forms. The degree of grammaticalization between the main verb and the auxiliary of future was also analyzed according with criteria determined by Bybee, Pagliuca e Perkins (1991). The results show that the future indicative and conditional tenses in MP must be regarded as analytical forms, formed with whether the infinitive of the main verb and verb aver conjugated in the present indicative, in the case of future indicative, or the infinitive of the main verb and verb aver/ir conjugated in the past imperfect indicative, in the case of the conditional. The fact that the parts forming these verbs could be written separately (and even invertedly in some cases), and also that a preposition... (Complete abstract click electronic access below)

Wernicke’s aphasia (WA), an acquired impairment of language comprehension and word repetition, results from a cerebrovascular accident to the left temporoparietal junction. The disorder has been important to the development of neurobiological models of language, however neuropsychological investigations into the nature of the comprehension impairment have been limited. This thesis presents a series of four experiments, investigating the comprehension impairment in WA. Chapter 3, a behavioural neuropsychological study, investigates existing hypotheses of the comprehension impairment in WA: a phonological breakdown, a semantic breakdown, a dual phonological-semantic breakdown. A case series comparison methodology is utilised. Participants with WA are compared to participants from two other clinical, comprehension impaired groups: semantic dementia and semantic aphasia. Semantic dementia and semantic aphasia provide neuropsychological models of semantic breakdown, affecting semantic representations and semantic control respectively. Individuals with WA showed disrupted non-verbal semantic analysis of a similar magnitude to that in semantic dementia and semantic aphasia and of a qualitatively similar nature to that in semantic aphasia. A significantly greater impairment on assessments which required acoustic-phonological analysis was found for individuals with WA compared to semantic aphasia. Overall a dual breakdown in acoustic-phonological and semantic control best accounted for the comprehension impairment in WA. In Chapter 4, direct evidence was sought for a link between acoustic-phonological non-word analysis and auditory comprehension in WA. A novel test of non-word discrimination was created which was perceptually graded so as to provide a sensitive measure in severely impaired participants. Individuals with WA were significantly impaired at non-word discrimination compared to age and hearing matched control participants who performed at ceiling. The degree of non-word discrimination/acoustic-phonological analysis impairment correlated with auditory comprehension in WA. Chapter 5 investigated the extent to which the established acoustic-phonological impairment in WA was grounded in a more fundamental deficit in non-verbal auditory analysis. The capacity to detect structural changes in non-verbal auditory stimuli was measured. Participants with WA had an impaired capacity to detect differences in all but the most structurally simple auditory stimuli, compared to control participants. The degree of this impairment correlated with the degree of auditory comprehension impairment in the WA group. Chapter 6 revisits the semantic impairment observed in WA. Functional magnetic resonance imaging was used to investigate the residual neural networks recruited by individuals with WA, when performing a semantic animate-inanimate judgment task. Large portions of the inferior and anterior temporal lobes bilaterally were activated, regions remote from the lesion in WA. Age matched control participants recruited similar regions; however the activation in WA participants was significantly stronger. This indicated greater reliance on the residual semantic network in WA in response to damage to posterior temporoparietal semantic regions. The results from this series of studies indicated that the primary deficit in WA is one of impaired acoustic analysis and co-morbid damage to a phonological system. Additional disruption occurs to the semantic control network, which regulates the task directed use of semantic representations. A combination of all three factors accounts for the comprehension impairment in WA and it is the relative contributions of each factor that accounts for behavioural variation between individuals.

Cette description grammaticale de l’afar s’appuie essentiellement sur des données recueillies sur le terrain auprès de très nombreux locuteurs afarophones, vivant à Djibouti, en Erythrée et en Ethiopie. La richesse des données a permis de prendre en compte les variantes dialectales tout au long de la description. Celle-ci s’organise autour de plusieurs parties : système phonologique et phonétique, système nominal, système verbal, et syntaxe de l’énoncé. Cette étude a permis de mieux évaluer l’originalité de certains traits connus mais aussi de mettre en valeur des traits peu connus voire ignorés jusqu’à présent. Citons entre autres : redéfinition du statut de la rétroflexe ; approfondissement des valeurs sémantiques et syntaxiques attachées aux différents procédés de dérivation nominale et verbale ; illustration du rôle primordial de la relative dans une langue où il n’y a pas de catégorie « adjectif » ; étude détaillée de la polarité de genre et de l’accord du verbe en genre non en nombre (accord au singulier avec un nom marqué comme pluriel) ; distinction entre noms adverbiaux et adverbes sur des critères syntaxiques; mise en valeur de la portée syntaxique et sémantique des quatre postpositions; meilleure définition de la catégorie des idéophones sur le plan morphophonologique, syntaxique et sémantique ; mise en lumière des procédés de topicalisation. En apportant un nouvel éclairage sur l’afar, à l’intérieur du couchitique et de la famille afro-asiatique, cette thèse se veut aussi une contribution à la typologie des langues. / This grammatical description of Afar is mostly based on fieldwork data collected from a great number of Afar speakers living in Djibouti, Eritrea and Ethiopia. The wealth of data has allowed us to embrace dialectal variations in the description. This description has been built around several components: the phonological and phonetic system, the nominal system, the verbal system, and the syntax.This study leads us to better assess the originality of some features that are already known, and also to bring out other features that were not well known or unknown until now: redefinition of the status of retroflex consonants ; enhanced understanding of semantic and syntactic values of different processes of nominal and verbal derivation ; illustration of the crucial role of the relative clause in a language without adjectives ; detailed study of gender polarity and verb agreement in gender – and not in number – (agreement in the singular with a name marked in the plural) ; differentiation between adverbial names and adverbs on the basis of syntactic criteria ; highlight of the syntactic and semantic significance of four postpositions ; better definition of ideophones from the morphophonological, syntactic and semantic aspects ; highlight of topicalization processes. By shedding new light on Afar within the Cushitic and Afro-Asian family, this thesis aims also to contribute to language typolog*y.