Early Verb Marking in Mandarin-speaking toddlers

• In Event: 3-037 - Syntactic Development in Mandarin-Speaking Children

Integrative Statement

Few studies have examined verb marking in native speakers of Mandarin even though verbs represent a large portion of Mandarin-speaking children’s early vocabularies (Tardif, et al., 2008a). Moreover, extant research has found differing results, depending on the measure adopted, and few results combine measures of grammar with vocabulary and MLU in normative samples.

The current study examines the development of a comprehensive set of verb markers in 260 Mandarin-speaking children (aged 20 to 33 months, 143 boys and 117 girls) who were at the very earliest stage of producing complex sentences. For each child, a caregiver was asked to fill out the grammatical complexity questionnaire from the Chinese Communicative Development Inventories (Tardif et al., 2008b). Items were phrased in the form of “Which sentence most closely corresponds to how your child would express ….” and included items for Modality, Resultative Verb Complements (RVCs), Negation (mei2, bie2, bu4), the Perfective Aspect marker le and Experiential Aspect marker guo4.

Age was significantly correlated with use for each verb marker -- Modals (r = .38, t (258) = 6.6, p < .001), RVCs (r = .23, t(258) = 3.75, p < .001), Negation markers-mei2, bie2 and bu4 (r = .27, t(258) = 4.48, p < .001; r = .18, t (258) = 2.9, p < .005; and r = .24, t(258) = 4.03, p < .001, respectively), Aspect markers le and guo4 (r = .34, t(258) = 5.72, p < .001 and r = .28, t (258) = 4.63, p < .001, respectively).

Figure 1 shows that children 20 to 33 months were reported to use negative markers bu4 and mei2 in their speech frequently and these two markers are used by 95% of typically developing children by 27 months of age. Aspect markers, RVCs and modals (Figure 2) show a slower pattern of development, along with the negative imperative marker, bie2. Results from mixed linear regression models with random intercepts showed that the intercept of mei2 (b0(mei) = .19) was significantly larger than that of bie2 (b0(bie-mei) = - .29, p < .05) and smaller than bu4 (b0(bu-mei) = .35, p < .05). Similarly, the intercepts of RVCs and le are significantly higher than modals and guo4, b0(le-modal) = .31, p < .05, b0(RVC-modal) = .74, p < .001, b0(guo-modal) = -.08, p = .56. However, the slope for modals is larger than RVCs, b1(RVC-modal) = -.02, p < .001, and are produced by over 75% of typically developing children by 29 months (2;5).

Finally, and similar to English, although age is highly correlated with children’s production of verb markers, children's MLU, total vocabulary and verb-specific vocabularies are more strongly related to Mandarin-speaking toddlers development of syntactic markings. These data have important implications when testing children with suspected language delays. Normative data combining verb marking with vocabulary and MLU, especially across a diverse range of languages, allows researchers and clinicians to more carefully diagnose issues and provide a fuller theoretical perspective on how grammatical markers and vocabulary work together in language development.

Authors

• Twila Tardif, University of Michigan
  Non-Presenting Author

• Lexie Ruoyu Huang, University of Michigan
  Presenting Author

• Paul Fletcher, Cork University
  Non-Presenting Author

• Weilan Liang, Peking University First Hospital
  Non-Presenting Author

• Zhixiang Zhang, Peking University First Hospital
  Non-Presenting Author