Language Choices and Code-Switching by Neurotypical and Neurodiverse Bilingual Children in Different Language Environments

• In Event: Language and Literacy Outcomes of Neurotypical and Neurodiverse Children Growing up in Bilingual Environments

Abstract

Introduction: The Adaptive Control Hypothesis (Green & Abutalebi, 2013) describes three interactional contexts: single-language, dual-language, and dense code-switching. The dense code-switching context, which allows speakers to use all their linguistic resources opportunistically (e.g., Beatty-Martinez et al., 2020), may assist self-expression in children with disabilities who have limited resources in each language and/or potential executive function differences. However, code-switching studies of children with Developmental Language Disorder (DLD) usually have an expected target language (e.g., Gross & Kaushanskaya, 2022; Gutierrez-Clellen et al., 2009; Kapantzoglou et al., 2021), and there is limited work on code-switching by autistic children (Yu, 2016). The current study examined language choices of bilingual children (neurotypical, autistic, and with DLD) in a dense code-switching environment and differences from a single-language environment.

Hypotheses: The ability to modulate code-switching depending on the environment may require high bilingual proficiency. For children with limited skills in one or both languages (including those with DLD), code-switching may be a necessary resource for expression, and we may expect fewer differences between environments. We did not have a specific hypothesis for autistic children, given the paucity of research on code-switching and mixed findings about how autistic children adjust their register for different listeners (e.g., Volden et al., 2007).

Study Population: Data collection is in progress with children ages 4-6 who were exposed to Spanish from birth and can produce simple phrases in both Spanish and English. Preliminary analyses included 4 autistic children (3 boys, 1 girl), 8 children with DLD (5 boys, 3 girls), and 10 neurotypical children (5 boys, 5 girls). The majority of children were exposed to mostly Spanish at home and English at school, with varying exposure to code-switching.

Method: In a scripted confederate dialogue task, children took turns describing pictures with a partner in three language environments: with a bilingual partner who code-switched frequently, with a Spanish-speaking monolingual partner, and with an English-speaking monolingual partner. The partners were presented as pre-recorded videos over Zoom but appeared to be interacting in real time. Caregivers provided information about their child’s language environment, and children completed measures of language and cognitive skills.

Results: Individual data for use of English, Spanish, and code-switching in each environment are shown in Figure 1, and group box plots for code-switching (the outcome of interest) are shown in Figure 2. Autistic children exhibited limited code-switching, except one child who code-switched frequently in only the code-switching environment. For DLD and neurotypical groups, a mixed-effects logistic regression revealed that children code-switched more in the code-switching than the English environment (z=4.82), with no difference between Spanish and code-switching environments (z=1.08). There was no effect of group (z=0.65) but a tendency (z= -1.70) for the DLD group to show a smaller difference between the code-switching and English environments compared to neurotypical peers. Analyses with the full sample will examine moderators of this environment effect, such as relative proficiency and code-switching exposure. Findings will be contextualized within a developmental approach to the Adaptive Control Hypothesis and implications for incorporating code-switching into bilingual assessment and intervention.

Author

• Megan C. Gross, University of Massachusetts Amherst
  Presenting Author