Null Subject Occurrence in Monolingual Spanish SLI: A Discriminant Function Analysis

Abstract

Background: Child Spanish-speakers appear to use more null subjects than do adults. Null subject use, like the use of tense marking, is sensitive to discourse-pragmatics. Because tense marking has been used to identify child Spanish-speakers with specific language impairment (SLI) with near good sensitivity and specificity (89%), null subject use may as well, following the predictions of the Interface Deficit Hypothesis. We investigate the possibility that null subject occurrence may form part of a useful discriminant function for the identification of monolingual child Spanish-speakers diagnosed with specific language impairment. Methods: We evaluate the rate of null subject expression from spontaneous production data, together with results from independent measures of another discourse-sensitive construction, verb finiteness, in child Spanish. We perform a discriminant function analysis, using null subject expression as a target variable, among others, to classify monolingual child Spanish-speakers (N = 40) as SLI or as typically-developing (TD). Results: The SLI group is shown to have significantly higher scores than the TD group on null subject expression. Multiple discriminant functions, including the null subject variable with tense measures, and in combination with mean length of utterance in words (MLUw), are shown to provide good sensitivity and specificity (<90%) in the classification of children as SLI vs. TD. Conclusion: Our findings support the contention that null subject occurrence is a plausible reflection of the Interface Deficit of SLI for Spanish-speaking children.

Keywords

• Spanish
• null subjects
• SLI
• Interface Deficit
• Interface Delay
• discriminant function analysis