Diagnóstico dos mecanismos disfluentes em orações relativas do português brasileiro

Abstract

Some studies in Psycholinguistics area deal with the interruption of the natural form of fluency speech. These suspensions are called disfluencies and can be of various types, for example, repetitions, substitutions or pauses. Among some works about disfluency, the one made by Jaeger (2005) deserves attention. His research dealt with the omission of the relativizer that in English and the possibility of this factor being the responsible for the creation of disfluencies. With this in mind, we aimed to analyze disfluencies in the layer of relativizers - CP layer - in Brazilian Portuguese. For this, data were taken by Côrrea et al (2008) on production of relative clauses. In this study, the authors, trying to verify the complexity of the formulation of relative clauses, manipulated two conditions: totally planned which would reduce the processing load and, consequently, major production of standard relative clauses and partially planned which would increase the processing load and, therefore, major production of gap and resumptive relative clauses. So, knowing that the planning of more complex structures can be a factor that generates more disffluent sentences, the question in which of the conditions stipulated by Corrêa et al (2008), we would find major cases of disfluency. We believe that the condition partly planned results in more cases of disfluencies in comparison with the condition entirely planned. To investigate this hypothesis, this research has focused its attention on the database of Corrêa et al (2008). After certain criterions, were privileged sentences of genitive relative clauses, functional indirect object of relative clauses and lexical indirect object of relative clauses, because these would be syntactic constructions more complex and likely to have more cases of disfluencies. Once identified and quantified all kinds of speech that interested us, there were created tables in order to enable a better visualization of disfluencies in specific locations in sentences, for instance, before the relative clause (BRC), in relative clause (IRC) and after the relative clause (ARC). With such data analyzed, it was found that the data corroborate the hypothesis: in the partially planned condition there are more cases of disfluencies than there are in the totally planned condition, principally when the location IRC is focused. Among the main results, we noticed that the disfluencies such as filled pauses appear in large amounts in IRC location, a factor that reveals a special feature of the relative clause. This disfluency, in this stretch of the sentence, reveals that the speakers are not only looking for a lexical, but also a discursive planning. These results motivated us to think of some other research situations, for example, analysis of disfluencies in the passive voice sentences