The Effect of Animacy and Schooling in the Production of Co-referential Objects in the Acquisition of two Varieties of Portuguese

Description

This study focuses on the production of 3rd person anaphoric forms as direct objects by native-speaking children of Brazilian Portuguese (BP) in contrast with native-speaking children of European Portuguese (EP). In both varieties, these co-referential forms can be definite full DPs, pronominals and null forms, though there are differences between them in the type of pronoun and null object. Unlike EP, in which object pronouns are clitics, stressed pronouns predominate over the accusative clitics in oral, informal BP. The acquisition of clitics in this variety seems to be largely promoted by the late exposure to written language, still influenced by standard EP. BP and EP also differ in the status of null objects as bound variables or pro forms (although this is subject to debate). Hence, the conditions that determine the selection of lexical and null forms may differ in these two varieties of Portuguese. In EP, young children tend to omit clitics even in ungrammatical contexts (such as strong island contexts). A number of micro-parameters seems to guide the balance between clitics and null forms in the adult language (1,2,3). In BP, spontaneous production data suggest that animacy affects the distribution of null and stressed pronouns, null forms being preferred with [-animated] antecedents [4,5,6]. It is not clear whether restrictions might apply in island contexts since the status of null forms is not consensually established in the adult language [7, 8]. It is also unclear whether EP-speaking children are sensitive to animacy. The aims of this study are (i) to assess the effect of animacy in the use of null/stressed forms (BP) and null/clitics (EP); (ii) to compare the use of null forms in BP and EP in an island context; (iii) to assess the effect of age/schooling in the emergence of clitics. Elicited production experiments were conducted with BP-speaking children (4-15 years of age) and replicated with a smaller sample of EP-speaking children (6-15 years of age). Two tasks were used: a WH-question elicited the production of co-referential forms in simple sentences (task 1); a cloze task (task 2), the production of these forms in an island context (reason clause). The independent variables were animacy and age/schooling. In the BP-experiment, participants were preschoolers/literacy-class (aged 4-7); 1st segment (aged 8-11) and 2nd segment (aged 12-15) of Elementary School (ES). In the EP-experiment, participants aged 6-9 in the 1st, and 10-15 in the 2nd segment of ES. The distribution of stressed pronouns, clitics, null forms and full DPs in each variety is presented. The rate of null forms and pronouns (clitics in EP, and stressed forms in BP) were taken as dependent variables. Participants were 108 native speakers of BP, from the public schools in Rio de Janeiro (54 females, without cognitive/language dysfunctions), divided into 3 groups (36 participants each), and 20 native speakers of EP from the public education in the Lisbon area (7 females, without cognitive/language dysfunctions), divided into 2 groups (10 participants each). The material for each test consisted of 4 pre-test sentences, 2 introducing each type of task, and 12 experimental sentences. 28 pairs of Power Point slides were presented (4 for pre-test and 12 for each test). One slide presented the referents and the other, the event to be described in the child's response. In both tasks, an existential sentence introduced the relevant referents (E.g. Here's a boy and a cart). In task 1, a WH-question was formulated (e.g. What did the boy do with the cart?). In task 2, a simple sentence described the event (e.g. Look, the girl is adorning the ballerina), and a complex one, with a reason clause, should be completed by the child: (Ex. Now the dancer will look beautiful because the girl...). The two tasks were performed in the same experimental session. The experiment was conducted in isolated school rooms and took 5 minutes on average per child. The responses for each dependent variable were submitted to ANOVAs (factorial design) and to non-parametric tests. In BP, the effect of animacy was obtained on both null and stressed forms in the predicted direction in both tasks without restriction to
null objects in the island context. No effect of schooling was obtained on these forms. The number of clitics did not allow statistical analysis, but they occurred in the most advanced ES segments. Regarding EP, inanimate antecedents favored null objects, in task 1, mainly in the youngest group. In task 2, null objects were rare. Animacy did not significantly affect accusative clitics in task 1. There was a marginally significant effect in task 2, favoring inanimate antecedents. Dative clitics (as direct object) were also identified particularly with animate antecedents. These results suggest that at least from age 4, the differences between the two Portuguese varieties regarding the encoding of co-referential objects can be attested.

Period	29 Jan 2021
Event title	Null objects from a crosslinguistic and developmental perspective
Event type	Conference
Location	PortugalShow on map
Degree of Recognition	International