I observe that in an earlystage of child Catalan and Spanish, no overt subjects are used. At this same age and MEAN LENGTH OF UTTERANCE (MLU), child speakers of overt subject languages such as French, German, Dutch, and English use at least some overt subjects optionally. I explain this crosslinguistic variation bysuggesting that the adult target grammars vary with respect to the position in which overt subjects are realized. In the overt subject languages, subjects are realized in the canonical specifier-of-IP position, whereas in the null subject languages (such as Catalan and Spanish), subjects are located in a topic/focus position, which becomes accessible onlylater in development. As evidence for this, I show that overt subjects, fronted objects, and WH-questions begin to be used at the same point in development in child Catalan and Spanish. I also argue that subject agreement constitutes an incorporated pronominal subject in Catalan and Spanish and that children converge on this parametric option veryearly . The inabilityof child Spanish- and Catalan-speakers to use discourse-pragmatic information is explained as a delayin the development of the interface between grammar and discourse-pragmatics.

No abstract

The findings support the contention that the tense-marking deficit is a plausible clinical marker of SLI for Spanish-speaking children.

Evidence from child language development supports the position that overt subjects in Southern Romance languages are left-peripheral, Topic-Focus constituents. Specifically, overt subjects begin to be used at the same time as other less controversially left-peripheral, Topic-Focus constituents, such as fronted objects and wh-questions. However, this interpretation of the data would be much more compelling if it could be shown that these constituents do not emerge at the same time in the speech of children learning languages in which overt subjects are obligatory and largely independent of discourse considerations, such as German. To this end, we examined the speech of three longitudinally studied Spanish-speaking children, Carlos, Eduardo and Graciela, and the speech of two longitudinally studied German-speaking children, Simone and Caroline. We also examined the German data of Katrin and Andreas, which was collected on a single day for each. While the Spanish-speaking children begin productively using overt subjects, fronted objects and wh-questions at a statistically similar point in development, the German-speaking children's data show that overt subjects begin to be used significantly earlier than do fronted elements and wh-questions. This supports the argument that child Southern Romance and child German are different with respect to the timing of the development of these constituents. We believe that this reflects the children's early knowledge of the structure of their target adult languages, which in the case of Southern Romance, includes left-peripheral, discourse-sensitive subjects.

In this study, we investigate whether specific language impairment (SLI) manifests itself grammatically in the same way in Spanish and English with respect to nominal plural marking. English-speaking children with SLI are very proficient at marking plural on nouns. Spanish has two main nominal plural allomorphs: /s/ and /es/. The /es/ allomorph has received multiple theoretical treatments, including one (e.g., Harris (1991)) which argues that in singular-plural pairs such as florflores, the /e/ is epenthetic, while other accounts (e.g., Colina (2003)) argue that synchronically there is an underlying /e/ in the singular form (e.g., flore) which gets deleted by apocope. Child Spanish speakers with SLI in the United States have shown mixed results in their abilities to learn plural marking. They have shown low proficiency on an elicited production task, but have shown high proficiency in spontaneous production data. We show, using a new elicited production task in Mexico with a group of children diagnosed with SLI and two control groups, that performance is close to the high levels previously shown in spontaneous production studies. Further, we show that all children's performance with the epenthetic allomorph /es/ is worse than their performance with the canonical Correspondence should be sent to John Grinstead, 330 GRINSTEAD ET AL. allomorph /s/. Our results suggest that plural marking is not an axis of crosslinguistic variation between Spanish and English among children with SLI. On the basis of the absence of child errors of the flore type and presence of errors of the flors type, our data appear to support the epenthesis account of Harris (1991).

Existing research into speci c language impairment in Spanish utilizes primarily spontaneous production data and concludes that children do not have problems with verb niteness. In contrast, we show, through a new receptive measure, the "Grammaticality Choice Task", that the distribution of the 2 most common errors in child Spanish is su cient to distinguish language-impaired children from age and language control groups. We conclude that niteness marking on verbs is a promising clinical marker for SLI in Spanish and that spontaneous measures of niteness marking do not provide the most accurate representation of children's grammatical competence in the verbal domain in a null subject language such as Spanish.

No abstract

No abstract