| Resumo: | The goal of the present research was to investigate how the interpretation of clitic, strong and null pronouns by Portuguese preschool children is influenced by the grammatical status of those forms. In a first study, picture verification tasks were used in order to verify if the categorial status of object pronominal forms (clitic or strong) is intralinguistically relevant in European Portuguese (EP), especially in contexts of variation between non-reflexives and reflexives. According to the results, children did not have major difficulties with reflexive forms (anaphors), regardless of their clitic or strong status. In the interpretation of non-reflexive forms, their performance got close to the adults’ behavior with clitic pronouns, while it deviated with strong pronouns in prepositional contexts. Children overaccepted dispreferred coreferential readings when interpreting non-reflexive strong object pronouns in non-locative PPs. In a second study, truth value judgment tasks were applied with the intention of specifying if there is an interpretative asymmetry between null and overt pronominal subjects in indicative and subjunctive complement clauses. The results show that, in the indicative (with one or two intrasentential antecedents), children overaccepted the pragmatically inappropriate reading of coreference for overt strong subject pronouns, unlike adults. Children performed more adult-like with null subject pronouns in indicative clauses, when there is only one intrasentential antecedent (the matrix subject). However, they often accepted the dispreferred reading of disjoint reference with null pronominal subjects in the indicative, in the presence of two potential antecedents before the pronoun (the matrix subject and the matrix object). In the subjunctive (selected by volitional verbs or declarative verbs of order), children incorrectly assigned coreferential readings to both null and overt subject pronouns. Strong pronominal forms are argued to be licensed post-syntactically. The difficulties in the post-syntactic rejection of the dispreferred coreference when interpreting object and embedded subject strong pronouns (constrained by semantic and/or pragmatic factors) are based on processing problems at the interface level. Here, there is competition between convergent derivations and the comparison between those structures is costly for children’s limited working memory. In turn, clitic and null pronouns are licensed in syntax (making the establishment of the referential dependency of these forms to be more economical), since both are dependent on functional categories as inflection. However, there are some processing constraints in the interpretation of null pronominal subjects in indicative clauses, when the matrix object antecedent linearly intervenes in the referential dependency between the preferred matrix subject antecedent and the null embedded subject pronoun. In this case, children’s performance is guided by the linear proximity of the matrix object antecedent preceding the null pronoun. The subjunctive obviation (with both types of subject pronouns) is not completely acquired yet by children. Nevertheless, they show sensitivity to the contrast between the indicative and the subjunctive. The full mastery of obviation involves not only syntactic knowledge of the anaphoric nature of Tense (e.g. Meireles & Raposo, 1983) but also lexical and semantic knowledge of the matrix verbs, which takes some time to acquire. In the pronominal system, the more pronouns are syntactically licensed, the less problematic their acquisition becomes.
|
|---|