A qualidade de voz avaliada por meio de descritores impressionísticos e fonéticos: subsídios para o contexto forense em portal multimodal e multilíngue

Detalhes bibliográficos
Ano de defesa: 2024
Autor(a) principal: Scarpelli, Rafael Cury lattes
Orientador(a): Passetti, Renata Regina lattes
Banca de defesa: Não Informado pela instituição
Tipo de documento: Dissertação
Tipo de acesso: Acesso aberto
Idioma: por
Instituição de defesa: Pontifícia Universidade Católica de São Paulo
Programa de Pós-Graduação: Programa de Pós-Graduação em Linguística Aplicada e Estudos da Linguagem
Departamento: Faculdade de Filosofia, Comunicação, Letras e Artes
País: Brasil
Palavras-chave em Português:
Palavras-chave em Inglês:
Área do conhecimento CNPq:
Link de acesso: https://repositorio.pucsp.br/jspui/handle/handle/41206
Resumo: Comparing speakers in the forensic context is a complex task related to knowledge of speech sciences and technologies. An issue that arises in the forensic scenario is the identification of voices by lay witnesses. The description of voices by laypeople is impressionistic and commonly relies on the use of adjectives as semantic descriptors. However, there are voice assessment systems, such as Vocal Profile Analysis (VPA), which are successfully used to compare speakers. Experimental studies that include the pairing of impressionistic and phonetic descriptions have the potential to provide support for the qualification of expert reports. This work was developed within the scope of the multimodal/multilingual portal for the advancement of open science in the Humanities and aims to compare impressionistic assessments of voice qualities, based on semantic descriptors, with assessments using phonetic descriptors. These are research questions: what types of relationships are established between lay and technical terms? How are these terminological mappings organized? As research hypothesis, we assume that there will be pairings between impressionistic terms that refer to auditory sensations of vocal dynamics and quality. The theory that supports this research is the phonetic model for describing voice quality and the VPA. The perceptual evaluation of a set of voices was carried out in three stages: judgments by lay judges of the characteristics of the speakers' voices; judgments by voice experts based on qualifiers provided by lay judges; evaluation of voices by experts in the use of VPA. Pairings between technical and lay terms were made through multifactor analysis. Associations were found between VPA voice quality adjustments and adjectives used by listeners without training in phonetics to describe voices, such as the adjectives “weak”, associated with advanced tip blade and advanced tongue body, “strong”, associated with extensive mandibular range and extensive labial range, and “enthusiastic”, associated with lip spreading and tense vocal tract. This research project fills a gap in phonetic-forensic studies involving earwitness testimony because it contributes to the understanding of how voice quality characteristics impact listeners' perceptions