This dissertation challenges the standard view that fifteenth-century poets wrote irregular meters in artless imitation of Chaucer. On the contrary, I argue that Chaucer’s followers deliberately misread his meter in order to challenge his authority as a laureate. Rather than reproduce that meter, they reformed it, creating three distinct meters that vied for dominance in the first decades of the fifteenth century. In my analysis of 40,655 decasyllables written by poets other than Chaucer, I show that the fifteenth century was not the metrical wasteland so often depicted by editors and critics but an age of radical experimentation, nuance, and prosodic cunning. In Chapter One I present evidence against the two standard explanations for a fifteenth-century metrical collapse: cultural depression and linguistic instability. Chapter Two outlines an alternative framework to the statistical and linguistic methods that have come to dominate metrical studies. In their place I propose an interdisciplinary approach that combines the two techniques with cognitive science, using a reader-oriented, brain-based model of metrical competence to reframe irregular rhythms as problems that readers solve. Chapter Three applies this framework to Chaucer’s meter to show that the poets who inherited his long line exploited its soft structure in order to build competing meters; in that chapter I also argue that Chaucer did not write in iambic pentameter, as is generally assumed, but in a “footless” decasyllabic line modeled on the Italian endecasillibo. Chapter Four explores metrical reception; by probing scribal responses to Chaucer’s meter we can gain insight into how fifteenth-century readers heard it. Chapters Five through Seven investigate three specific acts of reception by poets: those of John Walton, Thomas Hoccleve, and John Lydgate. I conclude the dissertation by tracing the influence of Hoccleve and Lydgate on the later fifteenth-century poets George Ashby, Osbern Bokenham, and John Metham, and by identifying the eclipse of fifteenth-century meter with the Tudor poets Stephen Hawes and Alexander Barclay, who replaced a misreading of Chaucer’s meter with a misreading of Lydgate’s, inadvertently returning sixteenth-century poets to an alternating decasyllable reminiscent of Chaucer’s own meter. / text

La présente thèse étudie le rôle de l'accent prosodique (accent de mot vs. accent de syntagme) lors du traitement auditif de paires minimales ambigües (mots composés vs. syntagmes) du chinois mandarin. Deux types de paires minimales ont été utilisés: 1) Mots composés avec un ton neutre (ex: dong3xi0 « chose ») vs. Syntagme avec un ton plein (ex: dong3xi1 « est et ouest ») qui se distinguent par la réalisation du ton sur la syllabe finale ; 2) Mots composés Verbe-Nom (VN) (ex: 'chaofan « riz frit ») vs. Syntagmes Verbe-Objet (VO) (ex: chao'fan « frire du riz ») se distinguant par la position de l'accent prosodique. Nos données comportementales et neurophysiologiques démontrent que : 1) la syllabe finale est plus longue et l'étendue de la F0 est plus large dans les VO que dans les VN, 2) la prosodie assiste le système de traitement pour anticiper la structure morphologique des séquences ambigües, et 3) un traitement hiérarchique « de droite-à-gauche » des informations prosodiques en complément d'un traitement séquentiel « de gauche-à-droite » prend place en chinois mandarin. Prises dans leur ensemble, nos données précisent la description fonctionnelle et structurale du modèle Prosody-Assisted-Processing (PAP) pour le chinois mandarin. / The present thesis investigates the role of prosodic stress (i.e. lexical versus phrasal stress) on the auditory processing of Mandarin Chinese ambiguous compound /phrase minimal pairs. Two types of compound/phrase minimal pairs were used: 1) Compound word with a neutral tone (e.g. dong3xi0 "thing") vs. phrase with a full tone (e.g. dong3xi1 "east and west") distinguished by the final syllable tone realization; 2) Verb-Noun (VN) compound word (e.g. 'chaofan "fried rice") and Verb-Object (VO) phrase (e.g. chao'fan "fry the rice") distinguished by the position of the prosodic stress. Combined behavioral and neurophysiological data demonstrate that 1) the final syllable was more lengthened and the F0 range was larger in VO than in VN, 2) prosodic structure does assist the processing system in anticipating morphological structure, and 3) a right-to-left hierarchical processing of prosodic information in addition to a sequential left-to-right one is involved during the processing of ambiguous spoken sequences in Mandarin Chinese. Taken together, our findings allowed us to precise the functional and structural description of the Prosody-Assisted-Processing (PAP) model for Mandarin Chinese.

Le but de cette thèse a été d'explorer certains des mécanismes relatifs à la perception de la prosodie linguistique. Cet intérêt provient du fait que la prosodie, qui est portée dans le signal acoustique par les indices d'intensité, durée et pitch, pourrait guider les nourrissons dans l'apprentissage des mots et de leur ordre dans les phrases. Nous avons donc étudié un mécanisme qui pourrait sous-tendre ces capacités prosodiques précoces : la loi iambo-trochaïque (iambic-trochaic law : ITL). L'ITL (Woodrow, 1909; Hayes, 1995; Nespor et al., 2008) est un des principes d'organisation de la perception auditive qui a été proposé comme ayant un rôle important non seulement dans la perception chez l'adulte, mais aussi pendant l'acquisition du langage chez le nourrisson. L'ITL postule une préférence générale pour le regroupement des éléments sonores (syllabes, notes musicales...) en paires, en fonction des variations d'intensité, de tonalité (pitch) ou de durée de ces éléments : les sons forts ou aigus marquent le début de groupes de pattern trochaïque (fort-faible ou aigu-grave), alors que les sons longs marquent la fin de groupes de pattern iambique (court-long). Toutefois, les langues diffèrent dans leur façon d'utiliser ces indices acoustiques pour marquer la proéminence prosodique dans les mots et dans les syntagmes. Par exemple, le français ne possède pas d'accent lexical mais un accent au niveau du syntagme où la dernière syllabe est allongée, ce qui crée un contraste court-long (pattern iambique basé sur la durée). A l'inverse, en anglais ou en allemand, l'accent lexical est en général sur la première syllabe, qui est plus forte et/ou plus aigüe, ce qui induit un pattern trochaïque basé sur l'intensité ou le pitch. L'environnement linguistique pourrait donc interagir avec l'ITL. Cette thèse s'articule en deux axes. D'une part, dans une tâche de segmentation / reconnaissance de paires de syllabes, nos données montrent que des sensibilités à l'ITL sont présentes chez l'adulte (Exp. 1) et le nourrisson de 7,5 mois (Exp. 2), français et allemands. De faibles différences cross-linguistiques entre les deux groupes sont retrouvées pour le groupement basé sur l'intensité. D'autre part, à l'aide de la NIRS (Near InfraRed Spectroscopy), nous avons mesuré les réponses d'activité cérébrale chez des nouveau-nés et mis en évidence que des sensibilités à l'ITL sont présentes à la naissance et déjà influencées par la langue entendue in utero (Exp. 3-6). Pour finir, nous avons exploré de quelle manière l'environnement linguistique influence les compétences de discrimination de patterns rythmiques d'accentuation des mots chez des nourrissons de 10 mois (Exp. 7). Nos résultats révèlent que des nourrissons bilingues apprenant le français et une langue avec un accent lexical variable sont capables de discriminer des patterns d'accentuation contrairement aux monolingues français. Pris ensemble, nos résultats contribuent à la compréhension de l'origine développementale de l'ITL, de sa modulation par l'environnement linguistique et du développement langue-spécifique du traitement et des préférences rythmiques. / The goal of this doctoral dissertation was to explore the mechanisms underlying linguistic prosodic perception. Prosody is carried in the speech signal by a number of acoustic cues, including duration, intensity and pitch. Importantly for language acquisition, prosody could help infants learn words and word order in their native language. Therefore, we studied a mechanism that could support these early prosodic abilities: the iambic/trochaic law (ITL). The ITL (Woodrow, 1909; Hayes, 1995; Nespor et al., 2008) is a mechanism that organizes auditory perception and was proposed to have an important role not only in adult speech perception but also in language acquisition in infancy. The ITL states that sounds (e.g., syllables, musical notes, etc.) contrasting in intensity/pitch form pairs with initial prominence, i.e., a trochaic pattern (strong-weak or high-low), and those contrasting in duration form pairs with final prominence, i.e., an iambic pattern (short-long). However, languages differ in how these acoustic cues mark prosodic prominence both at the level of words and of phonological phrases. For example, French has no lexical stress but has phrase-final stress, the last syllable of the phrase being lengthened, creating a short-long pattern (duration-based iambic pattern). Conversely, in English or in German, lexical stress is usually on the first syllable, which has higher intensity and/or pitch (intensity- or pitch-based trochaic pattern). Listeners' language background is therefore likely to interact with the ITL bias. This thesis is divided into two main parts. First, in a segmentation / syllable pair recognition task, we found that sensitivity to the ITL was present in French and German adults (Exp. 1) and 7.5-month-old infants (Exp. 2). We found weak cross-linguistic differences between the two language groups for the intensity grouping in adults and infants. Secondly, using NIRS (Near InfraRed Spectroscopy), we measured cortical responses in newborns and demonstrated that sensitivity to the ITL was present at birth and were already influenced by the language the infants heard in utero (Exp. 3-6). Finally, we observed how language background influences the ability to discriminate lexical stress patterns in 10-month-olds (Exp. 7). Our findings show that bilingual infants simultaneously learning French and a language with variable lexical stress were able to discriminate stress patterns whereas monolingual French infants could not. Taken together, our results contribute to a better understanding of the developmental origin of the ITL, its modulation by linguistic experience, and language-specific processing and rhythmic preferences.

Αντικείμενο της παρούσης διδακτορικής διατριβής αποτελεί η μελέτη και μοντελοποίηση των φαινομένων επιτονισμού της Ελληνικής γλώσσας με εφαρμογές στην σύνθεση ομιλίας. Στα πλαίσια της διατριβής αυτής αναπτύχθηκαν πόροι ομιλίας και εργαλεία για την επεξεργασία και μελέτη προσωδιακών παραγόντων οι οποίοι επηρεάζουν την πληροφορία που μεταφέρεται μέσω του προφορικού λόγου. Για την διαχείρηση και επεξεργασία των παραπάνω πόρων υλοποιήθηκε πλατφόρμα μετατροπής κειμένου σε ομιλία βασισμένη στην συνένωση δομικών μονάδων ομιλίας. Για την μελέτη και την δημιουργία των μοντέλων μηχανικής μάθησης χρησιμοποιήθηκε η γλωσσολογική αναπαράσταση GRToBI των φαινομένων επιτονισμού. / In this thesis we cope with the task of studying and modeling prosodic phenomena encountered in Greek language with applications to the task of speech synthesis from tex. Thus, spoken corpora with various levels of morphosyntactical and linguistic representation as well as tools for their processing, we constructed. For the task of coding the emerged prosodic phenomena of our recorded utterences we have utilized the GRToBI annotation of speech.

Dialektal medvetenhet är en förmåga som grundas i en metaspråklig förmåga, det vill säga att kunna resonera kring språk och hantera språkliga enheter. Exempel på detta är att med hjälp av prosodi kunna utläsa information i tal såsom känslor och dialekter. Det finns relativt lite forskning angående barns dialektala medvetenhet. Däremot finns mer forskning inom metaspråklig utveckling. En god metaspråklig förmåga hör samman med god generell språkutveckling och en metaspråklig kompetens innebär ett senare stadium av språkutveckling. Resultat från tidigare studier har visat att metaspråklig förmåga börjar utvecklas vid en ålder av cirka fem år vad gäller bland annat intonation och dialektrelaterat tal. I föreliggande studie konstruerades ett testbatteri för att kartlägga dialektal medvetenhet i åldrarna fem till tolv år. I studien deltog tre åldersgrupper, 5-, 8- och 11-åringar, med tio barn i respektive åldersgrupp. Testet är uppbyggt i tre delar som undersöker olika aspekter av dialektal medvetenhet. Studien är genomförd i Östergötland med barn som förväntas identifiera sig med östgötsk dialekt. Resultatet i föreliggande studie bekräftar att viss metaspråklig kunskap finns etablerad hos de femåriga barnen och att det sker en gradvis ökning i utvecklingen mellan åldrarna vad gäller att kunna avgöra om två personer låter som att de kommer från samma region. En liknande ökning av metaspråklig utveckling finns mellan respektive åldersgrupp och visas i barnens sätt att resonera kring olika sätt att tala. Föreliggande studies resultat visar även att barnen vid åtta år uppnått en god förmåga att urskilja sitt eget sätt att tala bland andra regionsdialekter, dock påvisas ingen ytterligare ökning av denna förmåga vid elva års ålder. / Awareness of accents is an ability derived from metalinguistic skills, which includes being able to reason about language and being able to handle small language units. An example of this may be to be able to use prosodic features in order to understand expressions of feelings and regional accents. There is not much research regarding regional accent awareness in Swedish children. There is, on the other hand, more research available about metalinguistic ability in children. Metalinguistic ability is associated with a more advanced language development in general, which means that metalinguistic capacity belongs to a latter phase of language development. Results from recent research have shown that metalinguistic ability starts to develop by the age of five regarding intonation and regional accents. The present study has produced a test battery to survey accent awareness in the ages five to twelve years. In the study participated three age groups of 5, 8 and 11 years with ten children from each age group. The test comprises three parts that investigate different aspects of awareness of Swedish regional accents. The study is carried out in Östergötland in Sweden, with children expected to identify themself with the Swedish regional accent Östgötska. The results of the present study confirm that certain metalinguistic knowledge is established by the age of five and that there is a gradual development between the age groups regarding the ability to decide if two persons sound like they are coming from the same region.  A similar gradual development of metalinguistic awareness in respective ages is shown in the children’s way to reason about different ways to talk. The results of the present study also indicate that children by an age of eight have achieved the ability to distinguish their own way of talking among other regional accents. No further development of this ability by the age of eleven is demonstrated.

Notre travail porte sur l’histoire de l’opéra vue à travers le double prisme du mythe et de la prosodie. Dans une première partie, nous définissons chacun de ces termes - mythe, prosodie - et envisageons ceux-ci dans les contextes précis de l’histoire de l’art, de l’histoire de la musique, enfin de l’histoire de l’opéra. Ainsi, l’aspect généalogique de notre travail est exposé dès cette première partie introductive, avant que chacune des parties suivantes ne soit consacrée à une, deux ou trois œuvres, siècle après siècle, de façon chronologique. Il nous a semblé important de rapprocher mythe et histoire pour aborder le XVIIe siècle qui voit tout d’abord la naissance de l’opéra liée à un certain retour au monde antique et à ses mythes, puis l’émergence, avec la dernière œuvre de Monteverdi notamment, de sujets historiques. Le XVIIIe siècle nous amène en France avec l’épanouissement de la tragédie lyrique, tandis que le XIXe siècle nous porte en Allemagne, sur les terres du romantisme. Le point d’aboutissement de notre parcours est l’opéra de Betsy Jolas, Schliemann, créé en 1995. / Our research deals with the history of opera as observed through the double prism of myth and prosody. We begin with a definition of these terms - myth, prosody - and examine them within the specific contexts of the history of art, the history of music and finally the history of the opera. The genealogical aspect of our work is thus presented forthwith at this introductory stage following which we shall, at each of the following stages, focus specifically on one, two or three works, century by century, in chronological order. It appeared important to us to bring together myth and history in order to deal with the 17th century which first witnessed the birth of the opera as a result of a return of sorts to the ancient world and its myths, and then saw the emergence of historical themes, notably illustrated by the final work of Monteverdi. The 18th century brings us to France with the tragédie lyrique in full bloom while the 19th century transports us to Germany on the footsteps of the romantic movement. Our journey culminates with the opera Schliemann by Betsy Jolas staged for the first time in 1995.

Children with Developmental Language Disorder (DLD) have difficulties in acquiring language in the absence of other neurodevelopmental issues (e.g. autism, hearing impairment) and despite growing up in an adequate language-learning environment. Previous characterisations of DLD have focused on grammatical processing, phonological memory or rapid auditory processing. This thesis approaches the language-learning difficulties of children with DLD from a novel perspective by considering the potential contribution made by differing levels of sensitivity to the rhythmic properties of language. Children with DLD have been shown to have reduced sensitivity to some of the acoustic cues present in speech which are thought to be important for rhythmic perception. Since rhythm forms the basis of language processing in early development, poorer sensitivity to language rhythm may result in later language problems. To investigate whether children with DLD demonstrate difficulties in processing language rhythm, this thesis explores five areas of language processing which could be affected by poor rhythmic sensitivity: locating word-boundaries, processing novel words, storing lexical stress patterns, representing sentence level structures and the integration of rhythm and syntax. As part of the investigation, measures were also taken of acoustic threshold sensitivity to see whether task performance related to acoustic sensitivity. A parallel strand of the study investigated whether provision of an entraining rhythm prior to task stimuli could support task performance. Three groups of children participated in the study: children with DLD, age-matched TD children (AMC) and younger, language-matched TD children (YLC). The results indicate that rhythmic manipulation of language stimuli affects task responses across the five language areas under investigation. The findings are then discussed in terms of the contribution made to our understanding of the role of rhythm in language and language disorder.

Dans une approche multi-théorique, cette étude propose d'observer des emplois en contexte des marqueurs genre en français, like en anglais et so en allemand, à travers des occurrences attestées et issues de productions (principalement orales) présentant un faible degré de planification. Cette analyse contrastive part du constat selon lequel ces trois termes présentent des emplois non-standard similaires à travers les langues, malgré des origines et des fonctionnements standard différents. Nous envisageons leurs caractéristiques communes du point de vue du changement linguistique qu'est la « grammaticalisation ». L'observation du corpus d'étude nous amène ensuite à analyser genre, like et so en convoquant différents outils théoriques syntaxiques et macro-syntaxiques, prosodiques et énonciatifs. Dans des configurations en « liste », telles que des énumérations, des reformulations ou après des segments interrompus de la production orale, nous examinons dans quelle mesure ces trois mots fonctionnent comme marqueurs de la construction progressive du discours. Cette étude se poursuit par l'analyse d'un fort point de congruence entre les trois marqueurs, à savoir l'introduction de discours direct. / Through a multi-theoretical approach, and using a corpus of authentic examples of unplanned conversations, this dissertation aims at describing the markers genre in French, like in English and so in German. This contrastive analysis is based on the observation that these three words display similar non-standard functions across languages despite their different origins and standard functions. Their common features are investigated from the theoretical perspective of « grammaticalization » as a type of linguistic change. Through the observation of the study corpus, we analyze genre, like and so with different theoretical tools, syntactic and macro-syntactic, prosodic and enunciative. In « list » configurations, such as enumerations, reformulations or interrupted segments of spoken discourse, we present how these three words work as markers of the progressive construction of discourse. This study further investigates the quotative function of genre, like and so as a strikingly congruent feature.

Made available in DSpace on 2014-06-11T19:25:38Z (GMT). No. of bitstreams: 0 Previous issue date: 2012-03-06Bitstream added on 2014-06-13T20:33:25Z : No. of bitstreams: 1 abreu_th_me_arafcl.pdf: 2188172 bytes, checksum: f6790b205dcb3b5c580af3b89b2373ac (MD5) / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) / Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP) / Esta Dissertação tem como objetivo principal analisar o estatuto prosódico das palavras aumentativas e diminutivas no Português Arcaico (PA) como formas simples ou compostas, a partir da observação dos fenômenos prosódicos desencadeados pelo processo morfofonológico da adjunção dos sufixos de grau -inno e variações, para o diminutivo, e –on(a), para o aumentativo, no galego-português, século XIII. As formas aumentativas e diminutivas em PA foram analisadas por meio de um corpus constituído pelas 420 cantigas religiosas em louvor à Virgem Maria, denominadas Cantigas de Santa Maria (CSM), compiladas pelo Rei Sábio de Castela, Afonso X (1221-1284), e pelas 431 cantigas de escárnio e maldizer, contidas nos cancioneiros profanos. A metodologia deste estudo é similar à proposta por Massini-Cagliari em seus trabalhos de 1995 e 2005 - por meio da escansão dos versos em que se encontram as ocorrências mapeadas pudemos localizar o acento poético e, consequentemente, o acento nas palavras, facilitando a investigação da estrutura prosódica dos nomes aumentativos e diminutivos no período arcaico de nossa língua. Após o mapeamento das ocorrências, foram investigadas algumas propriedades dos diminutivos e dos aumentativos sob a ótica da Fonologia Lexical (FL) e da Fonologia Prosódica. A partir disso, pudemos concluir que tanto as formas aumentativas como as formas diminutivas em -inn(o,a) são formas simples, uma vez que a Regra de Atribuição de Acento nessas formas ocorre em seu interior e, devido a isso, pode-se afirmar que os aumentativos e os diminutivos em -inn(o,a) correspondem a apenas uma palavra fonológica com um único acento, que é formada no léxico. Por outro lado, os diminutivos em -cinn(o,a) apresentam comportamento prosódico um tanto diferenciado dos aumentativos e diminutivos em -inn(o,a). Constatou-se que o padrão acentual destas formas aponta... / This dissertation aims to analyse the prosodic status of augmentative and diminutive words in Archaic Portuguese (AP) as simple or compound forms, through the observation of prosodic phenomena produced by the morphophonological process of grade suffixes attachment: -inno and its variations for the diminutive, and –on(a), for the augmentative, in Galician Portuguese, 13th century. The augmentative and diminutive forms have been analysed from a corpus composed by 420 religious cantigas in honor of Virgin Mary, called the Cantigas de Santa Maria (CSM), compiled by the wise king of Castile, Alfonso X (1221-1284), and 431 satirical cantigas, called “cantigas de escárnio e maldizer”, contained in the compilations of secular poems. The methodology is similar to that proposed by Massini-Cagliari in her studies (1995 and 2005): through the metrified verses where the mapped words appear we can find the poetic stress and, consequently, the word stress, making the investigation of prosodic structure of augmentative and diminutive forms in Archaic Portuguese easier. Thereafter the mapping of these forms some properties of diminutive and augmentative were investigated by way of Lexical Phonology (LP) and Prosodic Phonology. From this, we conclude that not only the augmentative but also the diminutive words in -inn(o) are simple forms, because the Stress Attribution Rule in these nouns occur in the limits of the word and, due to this, it is possible to state that augmentatives and diminutives in -inn(o) correspond to only one phonological word with only one stress, which is formed in the lexical level. On the other hand, the diminutives in -cinn(o) have a prosodic behavior rather differentiated. We noticed that the stress pattern of these forms point to the existence, like in Brazilian Portuguese (BP), of two prosodic prominences. Therefore, the Stress Attribution Rule is applied between... (Complete abstract click electronic access below)

Made available in DSpace on 2015-05-14T12:42:41Z (GMT). No. of bitstreams: 1 arquivototal.pdf: 3321436 bytes, checksum: bb2d715db22055d4290594233f8f8d75 (MD5) Previous issue date: 2011-12-21 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior - CAPES / This research has as main objective to understand the functioning of joint attention in mother-blind child interaction. For this, we conducted a longitudinal investigation, whose corpus was composed of videotaped records, with an average of fifteen to thirty minutes of occurred daily situations (free play, bath time, feeding, etc.) with that dyad in the home environment and a period of approximately one year. We selected eighteen out of these situations for review in order to monitor qualitative changes in scenes of joint attention established between mother and her blind son during this period. We based our study in view of the multimodal language functioning proposed by Kendon (1980, 2002, 2004), McNeill (1985, 1992, 2000), Butcher, Goldin-Meadow (2000) and Cavalcante (2009). In our theoretical path, we went on by contributions of joint attention, by discussions on the prosody and the roles of gestures in the mother-child interaction, and by a few descriptions of the effects of blindness in the process of joint attention. We started from the hypothesis that the scenes of joint attention are privileged moments to enter the child at the language; that such scenes are structured by multimodal language settings in the mother-blind child interaction, in which the voice/speech and touch are the main constituents of joint attention; that the prosody/motherese speech mother works as an access of blind children in these scenes; and the touch that has the status of the look in the operation of that child s joint attention. Our data showed that cases of child engagement in the interaction of joint attention with the mother contributed to the verbal acquisition of child; that this operation involved the integration of multimodal maternal and child elements, including maternal speech with prosodic markings and the touch were paramount in establishment of joint attention in the interaction with the child; in this process the voice quality in falsetto present in the motherese was the key to engage the child in this interaction, in which the child touch played an important role to replace the status of the look. These findings confirm our thesis that functioning in children with blindness occurs with the inclusion of touch and voice/speech, which serve as primary resources for their inclusion in the language. We hope these findings encourage reflection to think about alternative ways to engage the blind child in scenes of joint attention, contributing to language acquisition. / Esta pesquisa tem como objetivo principal compreender o funcionamento da atenção conjunta na interação mãe-criança cega. Para isso, realizamos uma investigação longitudinal, cujo corpus foi composto por registros videografados, com duração média de quinze a trinta minutos, de situações cotidianas (brincadeira livre, hora do banho, alimentação, etc.) ocorridas com essa díade no ambiente domiciliar por um período de aproximadamente um ano. Selecionamos dezoito dessas situações para a análise de forma a acompanhar mudanças qualitativas nas cenas de atenção conjunta estabelecidas entre mãe e filho cego. Fundamentamos nosso estudo na perspectiva do funcionamento linguístico multimodal proposta por Kendon (1980, 2002, 2004), McNeill (1985, 1992, 2000), Butcher; Goldin-Meadow (2000) e Cavalcante (2009). Em nossa trajetória teórica, percorremos contribuições acerca da atenção conjunta; por discussões sobre os papéis da prosódia e sobre os gestos na interação mãe-criança; e por escassas descrições sobre as repercussões da cegueira no processo de atenção conjunta. Partimos das hipóteses de que as cenas de atenção conjunta constituem momentos privilegiados para a entrada da criança na linguagem; de que tais cenas são estruturadas por configurações linguísticas multimodais na interação mãe-criança cega, na qual a voz/fala e o toque são os principais constituintes de atenção conjunta; de que a prosódia/manhês da fala materna funciona como via de acesso da criança cega nessas cenas; e de que o toque tem o estatuto do olhar no funcionamento da atenção conjunta dessa criança. Nossos dados mostraram que situações de engajamento na interação de atenção conjunta com a mãe contribuíram para aquisições verbais da criança; que esse funcionamento envolveu a integração de elementos multimodais maternos e os da criança; entre os quais a fala materna com marcações prosódicas e o toque foram primordiais na constituição da atenção conjunta na interação com a criança; neste processo a qualidade vocal em falsetto presente no manhês foi fundamental para engajar a criança nas interações, nas quais o toque teve papel de destaque ao substituir o estatuto do olhar. Esses achados confirmam nossa tese de que esse funcionamento na criança com cegueira ocorre com a inclusão do toque e da voz/fala, que servem como recursos primordiais para sua inserção na linguagem. Esperamos que essas constatações favoreçam reflexões para se pensar em caminhos alternativos para engajar a criança cega em cenas de atenção conjunta, contribuindo para a aquisição da linguagem.