Análise de processos semântico-cognitivos envolvidos em tarefas de desambiguação da referência pronominal

Abstract

Pronominal anaphoric ambiguity is a linguistic phenomenon caused when in an utterance two antecedent terms share morphological characteristics of person, number and gender with the personal anaphoric pronoun. This reality influences the anaphoric pronoun to seek its referential content in more than one lexical item that precedes it, generating ambiguity and causing a problem of Reading Comprehension. In order to investigate this problem, we applied two tests of pronominal anaphoric ambiguity resolution with students from the Youth and Adult Education - EJA, the first in compound sentences in the active voice and the second in simple sentences with two clauses in the active voice. was to understand the process of resolving pronominal anaphoric ambiguity being carried out by groups of different age groups. In the tests, we controlled the Independent Variables: position of the candidate for the anaphoric antecedent in the sentence and the Semantic Trait carried by the candidate for the anaphoric antecedent, in order to observe the influences of such variables on the Dependent Variable, that is, on the responses of the tasks of the tests. We chose to use the online experimental paradigm in order to analyze the importance of semantic information carried by a candidate for anaphoric antecedent for the process of resolving pronominal anaphoric ambiguity. We seek to observe whether this form of disambiguation generates more costs for memory, especially with the increase in the age of the reader. We performed the analysis of the data produced by the subjects in the two tests through the theories of: Antecedent Hypothesis, by Carminati (2002); Theory of Canonical Accessibility, by Schwenter (2003); Multifactorial Hypothesis, by Kaiser and Trueswell (2008); Distance Effect, by Streb et.al (2004) and Leitão and Simões (2011) and Animation Scale, by Comrie (1981) and Yamamoto (1999). The results relative to the general average suggest that the semantic disambiguation of the pronominal anaphor is a process that, with the increase of the reader's age group, sometimes takes more time, sometimes it takes less time to be performed, this compared to other forms of disambiguation, such as, for example, the syntactic. In this relationship, the way the reader performs the disambiguation, which is influenced by linguistic and extralinguistic factors, ends up impacting the processing time of the anaphor. Another finding of this study is that there are indications in the results that semantic information is an important linguistic clue, but not decisive for the realization of the disambiguation of anaphora. What is observed is that the effectiveness of this process is influenced by several linguistic and extralinguistic clues used together by the reader when deciding on the most suitable anaphoric antecedent to resolve an anaphoric ambiguity.