Left-behind experience and language proficiency predict narrative abilities in the home language of Kam-speaking minority children in China

Abstract

Introduction: Studies have documented that child experiences such as external/environmental factors as well as internal factors jointly affect acquisition outcomes in child language. Thus far, the findings have been heavily skewed toward Indo-European languages and children in the Western, educated, industrialized, rich and democratic (WEIRD) societies. By contrast, this study features an understudied minority language Kam, and a group of so-called left-behind children in China growing up in a unique social-communicative environment. Methods: Fifty-five bilingual children aged 5–9 acquiring Kam as home language were assessed using the Multilingual Assessment Instrument for Narratives (LITMUS MAIN). Twenty-three “two parents-left” children (mean age = 6;8, range: 5;0–9;2) remained in rural areas while both parents went to cities for employment, and they were raised by their grandparents. Thirty-two were “one parent-left” peers (mean age = 7;3, range: 5;0–9;3) who also resided in rural areas but were raised by one parent. Oral narrative texts were analysed for macrostructure based on story structure (SS), story complexity (SC) and internal state terms (IS). The study examined whether and how narrative production is predicted by internal factors such as chronological age and linguistic proficiency of a child and an external factor such as left-behind experience. Four measures were scored as outcome measures: SS, SC, IS type, IS token. Four measures were taken as predictors: chronological age, left-behind experience, scores in a lexical production task, and scores in a sentence repetition task tapping expressive morphosyntactic competence. Results: Results showed that left-behind experience consistently predicted all four outcome measures, where the “two parents-left” children scored significantly lower than their “one parent-left” peers. Expressive vocabulary scores predicted three measures: SS, SC, and IS Token. Expressive morphosyntactic scores predicted SS and SC. Age, by contrast, did not predict any outcome measure. Discussion: These findings suggested that being left-behind by both parents may be a negative prognostic indicator for the development and maintenance of heritage language abilities in ethnic minority children. We further discussed the conceptual significance of what it means for a child to be left-behind, by relating to more basic external factors in language development, including caregiver educational level, and amount of home language and literacy support by the caretakers.