Luegi, P. 1 , Costa, A. 1 & Maia, M. 2
1 Universidade de Lisboa
2 Universidade Federal do Rio de Janeiro
Ariel (1996) proposes that the form of an anaphoric expression signals the accessibility of its antecedent and so the more salient an antecedent is, the less marked will be the anaphoric expression referring to it.
Many authors (for instance, Costa et al., 1998, and Costa et al., 2004, for EP; Corrêa, 1998, and Melo & Maia, 2005, for BP; Carminatti, 2002, for Italian, Alonso-Ovalle et al., 2002, for Spanish)have already demonstrated that, in pro-Drop languages, the null pronoun is preferred to refer to Subject entities. However, in all these studies, the Subject was always the first referred entity. Accordingly to Gernsbacher (1998) first mentioned entities are very salient, regardless its syntactic status. So, in the referred studies, it is not clear if this preference may be attributed only to syntactic factors or if it was due to the combination of syntactic and order of mention information.
In this study we contrast the impact of syntactic function (Subject versus OBL, prepositional verb adjunct) and order of mention in pronoun resolution (see (1)-(2)). We used complex sentences where the second clause is an adverbial temporal one whose Subject is a pronoun (covert/overt) that must be identified by an antecedent in the previous clause.
In order to verify the preference in the identification of the ambiguous pronoun, we used the Visual World Paradigm. Eye movements of participants when viewing a picture (with two characters and the scenery of the action) are registered while listening to sentences like (1) and (2):
(1) O mecânico trabalhou com o engenheiro na oficina quando Ø/ele remodelou…
The technique [SUBJ] worked for the engineer [OBL] in the garage when Ø/he remodeled…
(2) Com o engenheiro trabalhou o mecânico na oficina quando Ø/ele remodelou…
For the engineer [OBL] worked the technique [SUBJ] in the garage when Ø/he remodeled…