Efeitos linguísticos e paralinguísticos na inferência dos sentidos indicados por (eu) acho que em entrevistas sociolinguísticas

Abstract

In Brazilian Portuguese, I think that is a construction grammaticalized as a polysemous epistemic modalizer which may indicate meanings of certainty, doubt, and uncertainty (GALVÃO, 1999; FREITAG, 2003; GONÇALVES, 2003). These meanings are inferred by the speakers in the context of interaction. According to descriptive studies, there are linguistic clues associated with each one of them such as the kind of complement introduced by I think that, the discursive topic, and the involvement of the speaker (direct, indirect) with what is said. However, paralinguistic clues, e.g., acoustic clues and facial expressions, may also help in the inference of those meanings. From the acoustic point of view, doubt and uncertainty are characterized by a high mean of fundamental frequency, higher values of intensity, duration, and presence of silent and filled pauses, which is the opposite for certainty (SILVA, 2008; OLIVEIRA, 2011; ANTUNES; AUBERGÉ; SASA, 2014; FERNANDES; ANTUNES, 2017). There are also gestural differences: Swerts et al (2003) suggested that certainty is characterized by a neutral facial expression, while uncertainty is characterized by the contraction in the lines of the eyebrows and of the mouth. When contracting facial muscles, speakers demonstrate expressions which, according to Ekman (2000), are important to keep the interaction. The aim of this study is to investigate the effect of linguistic, acoustic, and facial expression variables for the inference of the meanings of I think that in the sample “Deslocamentos 2020”, which is composed of 30 sociolinguistic interviews recorded on both audio and video, carried out with undergraduate students in the Federal University of Sergipe. We identified 1038 occurrences of I think that, which were codified in relation to meaning and analyzed in relation to linguistic (occurrence of I think that, scope, presence of modalizer, discursive topic, speakers’ experiences, and polarity), prosodic (means of the fundamental frequency, intensity, and duration of I think that, duration of silent and filled pauses) and facial expression variables (considering the movements of muscles that indicate anger, mockery, disgust, fear, happiness, neutral, sadness, and surprise, captured through a script in Python language). The meanings of I think that have association with the scope, presence of other modalizers, discursive topic, and speakers’ experiences, corroborating the results of other descriptive studies. For acoustic clues, variance analysis between the meaning of I think that and the prosodic variables suggested an effect only in the intensity and the duration. The meanings were differentiated by the strength of the sound and the extension of I think that: weaker and longer occurrences of this construction were characterized as doubt and uncertainty, respectively. The analysis of facial expressions evidences movements of contraction in the line of the eyebrows and of the mouth when the inferred meanings for I think that were of doubt and uncertainty, in contrast to the inferred meaning of certainty. The results of the analyses suggest that besides linguistic variables, paralinguistic resources are also relevant to the inference of the meanings of I think that, corroborating the perspective of language systematicity beyond the linguistic level.