Processamento da correferência anafórica de hiperônimos e hipônimos em português brasileiro: evidências de movimentação ocular

• Ano de defesa: 2018
• Autor(a) principal: Ferreira, Vitória Régia Santos
• Orientador(a): Soares, Maria Elias
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Palavras-chave em Português: Processamento Correferencial; Hiperonímia; Hiponímia; Frequência
• Link de acesso: http://www.repositorio.ufc.br/handle/riufc/47392

This dissertation presents a study in the area of Experimental Psycholinguistics focusing on the correference processing of hyperonyms and hyponyms in Brazilian Portuguese, based on the following theories: Centering Theory (GROSZ et al. 1983, 1995), The Theory of Accessibility (ARIEL, 1991), The Informational Load Hypothesis (ALMOR, 1999), recent studies in correference in Brazilian Portuguese (ALVES, 2012; LEITÃO, 2005; LEITÃO; QUEIROZ, 2005; SOARES; TEIXEIRA, 2014; TEIXEIRA, 2013), and findings from experimental research on frequency of words and eye tracking as an indication of that effect in on language processing (BRYSBAERT et al., 2017; SCHUESTER et al, 2016). We start from the hypothesis that in the context of Correference between hyperonyms and hyponyms, anaphoric resumptions, when represented by more frequent hyponyms, should be processed more quickly than less frequent words, and that the presence of the antecedent should ease the processing of less frequent anaphoric hyponyms. We used an online experimental methodology to prove our hypotheses with an experiment designed in four different conditions (2X2): (i) context without correference and with a more frequent hyponym, (ii) context without correference and with a less frequent hyponym, (iii) context with correference and with a more frequent hyponym, and (iv) context with correference and with a less frequent hyponym. Thirty-eight volunteer participants performed a silent reading of fortyeight phrases, the stimuli based on the four conditions, on a computer screen, while an eye tracer captured the participants' eye movement during the experiment. The dependent variables we used were: (i) the number of fixations; (ii) the time of first fixation; (iii) the average duration of fixation; and (iv) the total fixation time. Our results indicate that, during the establishment of the correferencing, there is no evidence of the presence of antecedent hyperonyms as a facilitating element of the cost of processing and of the frequency of the words as a property that interferes in the anaphoric resolution when the resumption is established by more frequent hygienic or less common.