Construindo categorias sonoras: o vozeamento de consoantes obstruintes em surdos profundos usuários de língua de sinais (LIBRAS)

• Ano de defesa: 2009
• Autor(a) principal: Rosana Passos
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Minas Gerais
• Palavras-chave em Português: Surdez; Fonética acústica; Lingüística; Fonética; Surdos Linguagem; Lingua portuguesa Consoantes; Sons da fala; Língua portuguesa Fonologia; Linguagem brasileira por sinais
• Link de acesso: https://hdl.handle.net/1843/ALDR-7RAKE8

This dissertation aims to analyse the voice properties of obstruent consonants in BrazilianPortuguese, specifically the stops [p, b, t, d, k, g], the fricatives [f, v, s, z, ., .] and theaffricates [t., d.], in pre-lingual deaf users of Brazilian Sign Language (Libras). The aim ofthe project was to investigate the contribution of dactylology (Libras manual alphabet) in theconstruction of sound categories by the deaf. Six deaf subjects and a control group of sixsubjects with hearing participated in five experiments. The experiments were composed oftasks designed to nominate words and logatomes in communicative speech and lip reading,dactylology and printed illustrations. The obstruent consonants investigated were located atthe initial position of the word. The deaf participants were between the ages nine to 14 andthose with hearing, between 18 and 30. All responses were recorded and filmed and the Praatsoftware program provided the acoustic analysis. The following acoustic measures wereinvestigated: the VOT (Voice Onset Time) of stops, duration of fricatives, VOT and durationof affricates, the presence/absence of bar voicing and the duration of the vowel following theobstruent. The results showed that the phonetic inventory of the deaf is reduced whencompared to those with hearing, being that all the deaf participants manifested only voicelessobstruents. The VOT values for stops were always positive, showing the presence of solelyunvoiced sounds. The absence of the bar voicing in the spectrogram confirmed that allobstruents investigated were unvoiced. VOT measures for stops, duration of fricative, VOTand duration of affricates presented an unsystematic characterization regarding the propertiesof voicing for obstruent consonants amongst deaf participants. It was not possible to find finephonetic detail generalizations which expressed the voice contrast in obstruents since all deafparticipants presented only voiceless sounds. Complementary research showed that thelengthening of the vowel which follows the obstruent occurred as a strategy to express thevoice contrast which was utilised by the deaf participants, principally by those who were moreproficient in sign language. It was furthermore observed that deaf participants who had ahigher degree of proficiency in Libras had a greater tendency to lengthen the vowels. The finephonetic detail in the construction of sound categories was investigated by measuring theduration of the vowel following the obstruent, according to the postulates of cognitivetheories of mental representation in the Usage-Based Phonology and Exemplar Models. Theresults indicate that the deaf use fine phonetic properties in the construction of specific soundcategories. The results also show that the categorization of voiced and voiceless sounds ismost effective when the deaf subject is exposed to various stimuli such as video, audio andLibras. The use of dactylology alone did not influence the construction of sound categories bydeaf participants in the tests. We suggest that these results express a holistic view of theconstruction of linguistic categories by the deaf. Future work will investigate the finephonological properties of other consonants, and also the role of Libras proficiency in theacquisition and use of specific phonological categories.