Diretrizes para a construção de Corpora Paralelos Libras-Português: uma proposta

• Ano de defesa: 2021
• Autor(a) principal: Malaquias, Keli Pereira
• Orientador(a): Bidarra, Jorge
• Banca de defesa: Bidarra, Jorge , Corbari, Alcione Tereza , Martins, Tania Aparecida , Santos, Patrícia Tuxi dos , Serpa, Talita
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Estadual do Oeste do Paraná; Cascavel
• Programa de Pós-Graduação: Programa de Pós-Graduação em Letras
• Departamento: Centro de Educação, Comunicação e Artes
• País: Brasil
• Palavras-chave em Português: Libras; Linguística de Corpus; Corpus paralelo Libras-Português; Diretrizes para corpus representativo
• Palavras-chave em Inglês: Libras; Corpus Linguistics; Parallel corpus Libras-Portuguese; Guidelines for representative corpus
• Área do conhecimento CNPq: Linguagem e Sociedade
• Link de acesso: https://tede.unioeste.br/handle/tede/5917

When considering the rise of linguistic research in Libras and the need for analysis of linguistic phenomena that occur in sign languages, this dissertation is the result of studies related to Corpus Linguistics (LC), since there is a significant increase of corpora involving these languages, both in Brazil and in other parts of the world. Thus, the main questions that arose were: (i) What would be the necessary criteria for the constitution of a parallel corpus between sign languages and oral languages?; (ii) What type of content should prevail in this corpus, based on the analyses to be performed?; and, more specifically, (iii) What needs to be considered in a parallel corpus, where, on one side, there are linguistic data from sign languages and, on the other side, the equivalent data from oral languages, from the perspective of the theories that underlie LC? In this context, the general objective of this research is to present and/or suggest, in a systematized way, the necessary guidelines for the construction of a parallel Libras-Portuguese corpus. For this purpose, the requirements were investigated, described and analyzed, as well as the types of information required for the constitution of a parallel corpus of sign language in interface with the spoken languages, from some corpora compiled or being compiled at national and international scales. For the analyses, we took into consideration not only the origin, the definitions and the concepts that the area has been referring to, but also the historical contexts involved and the characteristics necessary for a linguistic corpus to be, in fact, representative. In an attempt to meet the proposed objective and find answers to the problems raised, the following authors, among others, were used as theoretical support: Aijmer and Altenberg (2013), Aluísio and Almeida (2006), Anderman and Rogers (2008), Baker (1995), Beber Sardinha (1999; 2000; 2004), Granger and Petch-Tyson (2003), Kennedy (1998), McEnery (1996), Olohan (2004), Quadros (2006; 2008; 2011; 2016; 2017; 2018; 2019), Sanchez (2005), Sinclair (2005), Shepherd (2012), Tagnin (2018). Thus, the proposal of this work consisted of analysis materials, based on theoretical investigations and on corpora already constituted or under construction. It is a basic research, of a qualitative nature, supported by a bibliographic review work. At the end, some programs that can help in the construction of parallel corpora are presented and propose some guidelines that can contribute to studies that have been developed within the LC, notably with regard to the linguistic analysis of Libras in an interface to Portuguese.