Detalhes bibliográficos
| Ano de defesa: |
2013 |
| Autor(a) principal: |
Gonçalves, Cintia Schivinscki
 |
| Orientador(a): |
Brescancini, Cláudia Regina
|
| Banca de defesa: |
Não Informado pela instituição |
| Tipo de documento: |
Tese
|
| Tipo de acesso: |
Acesso aberto |
| Idioma: |
por |
| Instituição de defesa: |
Pontifícia Universidade Católica do Rio Grande do Sul
|
| Programa de Pós-Graduação: |
Programa de Pós-Graduação em Letras
|
| Departamento: |
Faculdade de Letras
|
| País: |
BR
|
| Palavras-chave em Português: |
|
| Área do conhecimento CNPq: |
|
| Link de acesso: |
http://tede2.pucrs.br/tede2/handle/tede/2115
|
Resumo: |
This study deals with the speech rate (SR) and the articulation rate (AR) in Brazilian Portuguese (BP) spontaneous speech obtained from unaware audio recordings (legally authorized telephone interceptions) and known ones (semistructured interview), situations commonly counterposed in the Forensic Phonetics area, specifically in Speaker Comparison (SC). The objective was to establish the speaker-discriminating power of SR and AR, contributing to the discussion about the effectiveness of their use as a comparative parameter in SC. In addition, the relationship between the rates and the independent variables of study (age, sex, formal education, temporal gap between the recordings, type of recording, length of speech stretch) was investigated. The SR and AR in the speech of seven subjects (five males and two females) was analyzed. The subjects live in the State of Rio Grande do Sul, Brazil, have BP as their L1 and undefined dialects due to occasional detention(s). The recordings belong to the database of Instituto Geral de Perícias, a state department of security. All subjects present a positive result for the SC. Measurement of the rates was carried out globally and locally. SRs calculation involved 539 turn talks, and ARs calculation involved 748 interpause stretches. Statistical analysis showed that there is no significant difference between the rate types (SR and AR) and between forms of measurement in SR. The AR was less variable than the SR, especially in the local measurement. As far as the speakerdiscriminating power of the rates, only the AR showed higher between-speaker variability than within-speaker variability, reaching a satisfactory intraclass correlation coefficient.The variability analysis by sex and type of recording was significant only for type of recording in SR. As a trend, it was observed that the subjects of both sexes decrease their rates when they are aware of the recording (reduction higher in males), and that in naturalistic speech (unaware recording) males predominantly presented the highest rates. Significant difference was found between the factors of the variable type of recording in SR and significant correlations between SR and length of speech stretch, and between AR and temporal gap. Based on the multiple local stretches, the significant predictors of SR and AR increase found in the analysis were sex (male factor), educational advancement, and higher temporal gaps. The significant predictors of decrease in SR and AR found in the analysis are advancing age and awareness of recording. It is possible to conclude that it is indicated to incorporate the AR local mean to the list of parameters used in the SC, provided there is a minimum temporal gap between contrasted recordings. Caution is recommended to minimize the impact of awareness of recording and of differences related to inherent speech style, as well as differences associated to educational advancement that might occur during the temporal gap between the confronted recordings |
