Linguistic Approaches to Bilingualism - Volume 12, Issue 4, 2022

The empirical evidence for whether narrative macrostructure skills are shared between a bilingual child’s two languages is inconclusive, and it is not known how macrostructure (overall story structure) is influenced by general language proficiency and amount of exposure. The present study investigates these issues in 100 Turkish-Swedish bilingual 4-to-7-year-old children growing up in Sweden. Oral narratives were elicited in both Turkish and Swedish with two picture-based tasks from the Multilingual Assessment Instrument for Narratives (MAIN) in the telling mode. We investigated to what extent the language of elicitation influences bilingual children’s macrostructure (story structure, episodic complexity), and explored effects of age, narrative task, narrative length, expressive vocabulary and language exposure, both separately and combined, on macrostructure in the respective language. Story structure and episodic complexity were found to increase similarly with age in both Turkish and Swedish from 4 to 7 years. Scores did not differ between the two MAIN storytelling tasks. Expressive vocabulary and narrative length influenced story structure scores positively and similarly in both languages. Daily language exposure and length of exposure to Swedish did not show any significant effect. The results can be interpreted in support of a carry-over of narrative macrostructural skills between the two languages.

Studies of macrostructure in bilingual children investigated potential age and language effects, without systematically taking into account bilingualism factors such as language dominance, exposure to narratives, and general language abilities. These issues were addressed in a study of macrostructure production by 69 bilingual Lebanese Arabic-French children aged 4–9. The children were administered the Multilingual Assessment Instrument for Narratives in the tell mode and a standardized conceptual vocabulary test. Parental questionnaires were also used to gather information on language exposure and use. Age effects and limited effects of language were found for all measures of macrostructure production (story structure, structural complexity and use of internal state terms). Significant correlations arose between these measures, vocabulary scores, and age, and between structural complexity scores and language dominance. Regression analyses revealed the predictive role of vocabulary in the development of macrostructure. These results are compatible with the idea that macrostructure is not language dependent. They also suggest that minimal language skills are needed for expressing macrostructure. The limited impact of bilingualism factors could stem from the fact that French and Lebanese Arabic are majority languages in Lebanon, meaning that the children had received continuous exposure to both languages, including within the school system.

Previous studies show mixed findings concerning whether higher-order story structure (macrostructure) is similar across bilinguals’ two languages. It is not known how macrostructure is influenced by general language proficiency and amount of exposure. The present study investigates these issues in 46 German-Swedish bilingual 4- to 6-year-olds. Narratives were elicited in both languages with two picture-based tasks from the Multilingual Assessment Instrument for Narratives (MAIN) in the telling mode. We investigate to what extent the language of elicitation (Swedish vs German) influences bilingual children’s macrostructure (story structure, episodic complexity) and explore effects of narrative task, age, narrative length, expressive vocabulary and estimated language exposure, both separately and combined, on macrostructure in the respective language. Results show that macrostructural skills developed measurably with age from 4 to 6 years in both languages, with no task effects. Story structure scores were higher in the majority language Swedish than in German and developed differently with age. The effect of narrative length on story structure was similar in the two languages. Language exposure did not have any significant effect. Macrostructure scores were significantly affected by expressive vocabulary in German only. Generally, the results may be linked to slightly higher language proficiency in Swedish.

The study explores the effect of language dominance on microstructure, macrostructure, and Internal State Terms (ISTs) in narratives of Russian-Hebrew bilingual children and examines within-language and cross-language associations between narrative elements in two dominance groups. Narratives were collected from 38 Russian-Hebrew bilingual children aged 5;5–6;7 using the LITMUS-MAIN retelling procedures. The children were divided into L1-dominant (N = 19) and L2-dominant (N = 19) bilinguals based on performance on proficiency tests in L1/Russian and L2/Hebrew. The narratives were coded for microstructure measures: number of different words (NDW), total number of tokens (TNT), number of C-units (CUs), and Mean Length of C-unit (MLCU); and for macrostructure measures: Story Structure and Story Complexity. Ratios of IST tokens and types were calculated per C-unit. Children produced significantly higher NDW, TNT, and MLCU in L2/Hebrew than in L1/Russian. Scores on macrostructure measures and ratios of total ISTs were similar across languages. L1-dominant bilinguals demonstrated similarity between L1 and L2 for microstructure and IST types, whereas L2-dominant bilinguals produced more IST types in L2/Hebrew and had relatively few significant cross-language correlations. Findings for language dominance and cross-language differences are discussed for those narrative features which emerged as sensitive to these effects.

There is no agreement regarding the relationship between narrative abilities in the two languages of a bilingual child. In this paper, we test the hypothesis that such cross-language relationships depend on age and language exposure by studying the narrative skills of 32 Indonesian-Dutch bilinguals (mean age: 8;5, range: 5;0–11;9). The narratives were elicited by means of the Multilingual Assessment Instrument for Narratives (MAIN) and analysed for story structure, episodic complexity and use of internal state terms (ISTs) in the home language (Indonesian) and majority language (Dutch). The results demonstrate that story structure scores in the home language (but not in the majority language) were positively related to age. Exposure measures (current Dutch/Indonesian input, current richness of Dutch/Indonesian input, and length of exposure to Dutch) did not predict the macrostructure scores. There was a significant positive cross-language relationship in story structure and episodic complexity, and this relationship became stronger as a function of length of exposure to Dutch. There was also a positive cross-lingual relation in IST use, but it became weaker with age. The results support the idea that narrative skills are transferable between languages and suggest that cross-language relationships may interact with age and exposure factors in differential ways.