Lo tengo, pero me cuesta: o processamento do clítico acusativo de terceira pessoa por bilíngues do par espanhol/português brasileiro

Abstract

The use of third person pronominal anaphora in the accusative case is a grammatical event that has divergent rules for Spanish and Brazilian Portuguese, as well as for the formal and informal registers of the latter, which are also called dialects by Guy and Zilles (2008), thePopular Brazilian Portuguese and the standard Brazilian Portuguese. The third person accusative clitic is considered to be a natural anaphor only in Spanish and in the standard register of Brazilian Portuguese. According to the multiple grammars theory (ROEPER, 1999; AMARAL; ROEPER, 2014), each of the divergent rule sets present in a speakersmind corresponds to a different subgrammar. The aim of this study was, then, to investigate whether processing of the third person accusative clitic by bilinguals proficient in languages or dialects which have different subgrammars concerning its usage would generate a linguistic conflict in the speakers mind, by reason of the fact that these rule subsetscompete. This prediction matches the Interface Hypothesis (SORACE; FILIACI, 2006; SORACE, 2011), according to which bilinguals face computational difficulties to integrate intralinguistic and extralinguistic sources of knownledge in the interface between syntax andpragmatics. This hypothesis was checked by confronting an online processing technique, the maze task, and an offline one, the acceptability judgment. If the linguistic conflict was present, it would be expected to find answers with different patterns between the two techniques (HOPP, 2009; SORACE; SERRATRICE, 2009). The judgment could be similaramong bilinguals and monolinguals to which the usage of the clitic is natural, that suggesting that the structure has a good acceptance rate. Or, alternatively, bilinguals judgment could achieve higher scores, as their greater metalinguistic awareness capacity (JESSNER, 2008) could turn them stricter to assess the structure adequacy. Regarding theonline task, it was expected that the reaction times obtained would be faster among bilinguals, in case in which there was a linguistic conflict. The findings of the research, with statistically and marginally significant results, pointed to the existence of linguistic conflict within the bilingual mind during the processing of the structure under analysis, as much tobilinguals who speak both Brazilian Portuguese and Spanish as to those who speak the standard and the popular dialects of this latter. The structure investigated has a high acceptance rate among these speakers in the judgment test, a fact that suggests that a linguistic attrition process is not occurring. On the other hand, the increased processingcharge observed in the online test could be explained by the cognitive charge of inhibiting the concurrent pragmatic rule, that is, of inhibiting the rule that did not correspond to the target language.