Individual difference factors and linguistic complexity in sequential bilingual acquisition

Abstract

Differences in bilingual language acquisition are often associated with variation in individual difference factors such as chronological age, quantity and quality of exposure, and cognitive capacities. Previous findings on the impact of these factors are mixed. However, factors are often confounded by other variables, tasks used, or divergent first language backgrounds. The present study aims to consider the impact of individual difference factors on sequential bilingual acquisition by taking a novel and multi-pronged approach to the analysis of issues that clearly demand a complex research paradigm. This includes incorporating the following measures: inclusion of a number of individual difference factors with separate measures of fine-grained input, analyses to reduce dimensionality between correlated variables; measurement of interdependencies between predicting factors; controlling the participants’ L1 background; measuring the impact of factors on both complex and non-complex language features; and the use of a novel experimental task comprising a simple and interactive format designed to reduce processing costs and age bias.

Participants were 40 Arabic-speaking children acquiring English as a second language (L2) (5;7 to 12;2 years old) alongside 40 age-matched monolingual controls. The dependent variable was accuracy of language measures: regular (third person singular) and irregular (past tense) morphology; canonical (actives) and non-canonical (passives) word order; subordinate clauses (relative clauses) and intervening elements in long distance dependencies (object relatives); and lexical proficiency. Independent variables were age of L2 onset, chronological age, short-term memory, motivation for learning the L2, richness of the L2 environment, L2 use at home, L2 use at school, and socio-economic status.

Generalised linear mixed effects modelling revealed the following for the bilingual participants: length of L2 exposure predicted vocabulary and morphology; chronological age predicted interpretation of sentences; short-term memory predicted production of morphosyntax; L2 richness predicted interpretation of all target sentences and interacted with chronological age for interpretation of complex syntax; and greater L2 use at home, a younger age of onset, and higher socio-economic status did not predict any measure. For the monolinguals, chronological age predicted vocabulary and morphology and production of morphosyntax while short-term memory predicted interpretation of syntax.

Results suggest the following for sequential bilingual children: overall exposure to language is important in the acquisition of vocabulary and morphology; greater maturation of cognitive and linguistic skills associated with older ages in childhood and greater participation in L2 play and social activities facilitate interpretation of sentence structures; increased participation in L2 play and social activities further enhances interpretation of complex sentences when children are older; short-term memory appears to play a role in processing expressive morphosyntax which may be linked to greater processing demands bilinguals experience with language phenomena when it interfaces with discourse; and starting to acquire a language at a younger age does not result in greater language proficiency. For the monolingual children, greater maturation of cognitive and linguistic skills associated with older ages in childhood facilitate most language features while short-term memory appears to play a role in processing the syntactic features of language.

This study builds on valuable previous research investigating individual differences in sequential bilingualism. By teasing apart the impact of factors, the study provides further insights on the effect of these factors across linguistic domains as well as novel findings on how predictors interact to impact language proficiency.

Date of Award	May 2022
Original language	English
Sponsors	Department of Education and Learning, Northern Ireland
Supervisor	Juliana Gerard (Supervisor), Christina Sevdali (Supervisor) & Raffaella Folli (Supervisor)