The benefits of multilingual pedagogies for multilingual children’s narrative abilities

ABSTRACT

The present study investigates whether exposure to multilingual pedagogies enhances emergent bilingual children’s narrative abilities. These abilities are among the most reliable indicators of children’s literacy skills. We compare two groups of emergent bilingual children with migrant background attending the fourth and fifth grade of a public primary school in Italy: 30 children were exposed to multilingual pedagogies, whereas 33 attended a ‘traditional’ monoglossic program. The results show that children exposed to multilingual pedagogies use more mental state terms in their narratives in both their home language and Italian than children exposed to monoglossic education. Furthermore children exposed to multilingual pedagogies exhibit similar patterns of use of mental state terms across the two languages. By contrast, the production of mental state terms differs across the two languages among the children exposed to monoglossic education. The qualitative analysis of the narratives shows that the home and the school language dynamically interact with each other regarding the construction of the narrative among the children exposed to multilingual pedagogies. The children attending monoglossic programs rely mostly on literal translations. The study shows that multilingual pedagogies boost emergent bilinguals’ literacy skills and enable them to access these skills from all languages in their repertoire.

KEYWORDS:

• Multilingual pedagogies
• multilingual resources
• metacognition
• narrative abilities
• mental state terms

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 All the authors contributed to data collection and analysis. Valentina Carbonara wrote sections 1.2, 2, 2.1, 3.3, 4.1, 4.3.1 and 4.3.2. Jacopo Torregrossa wrote sections 1, 1.1, 3.1, 3.2, 4.2 and 4.3.3. Andrea Scibetta wrote sections 2.2 and 4.3.4. Sections 5 and 6 were written together.

2 We use the following formula

m0 < – lm (event_structure_score ∼ cloze + DIFF_current_language_use + DIFF_home_language_history + DIFF_early_literacy + DIFF_current_literacy, data = narratives).

3 The resulting model is:

m0 < – lmer (event_structure_score ∼ 1 + group * language + (1|ID), data = narratives, control = lmerControl(calc.derivs = FALSE)). We did not include any random slop because of convergence issues.