Describing linguistic abilities in children entering primary school is relevant to setting appropriate expectations for development during the initial stages of schooling and advising concerned parents about the effects of early multilingualism. Since early differences in language skills may compound, creating increasing gaps in achievement, determining the best recipients for linguistic support early on is crucial for educators. In doing so, describing bilingual profiles (i.e., profiles of proficiency that consider both languages simultaneously) provides a valuable tool. This study investigates abilities in Catalan and Spanish at the beginning of Grade 1 in Catalonia, a province where Catalan and Spanish are co-official languages, where Catalan is the minority language.
Catalan-Spanish bilingual children (N=162) completed tests of receptive vocabulary and grammar in the two languages during the first two months of primary education (ages 5-6). Parents completed a background questionnaire that yielded information on participant/family demographics and their linguistic environment. We investigated two questions: 1-What bilingual profiles are present in the data? 2-What are the best predictors of these profiles?
Given previous research on bilingual development in Catalan-Spanish bilingual children, for question 1, we hypothesized that there would be two main profiles in the sample: one that is more Catalan-dominant (i.e., higher scores for the two Catalan tests and lower scores in the Spanish tests) and a Spanish-dominant one. For question 2, we hypothesized that the former group would be characterized by significantly higher exposure to Catalan in the home and at school, and vice versa for the Spanish-dominant group.
To answer question 1 empirically, we employed k-means clustering, which found two main bilingual profiles in the data. Contrary to our hypothesis, one bilingual profile could be described as high-performing in the two languages, with the second one appearing as relatively low-performing.
To answer question 2, we ran a logistic regression predicting whether a participant would be classified in the high-performing or in the low-performing group. The predictors were language use (in the home, with friends, and in the classroom), frequency of Catalan and Spanish use in reading/watching TV/extracurricular activities, and socioeconomic status. The model found two predictors that contributed significantly to the model. Specifically, it was found that participants in classrooms where Spanish was the most common language were significantly more likely to be classified in the low-performing cluster, and participants who engaged in reading activities more frequently in Catalan were more likely to be classified in the high-performing cluster.
Two main findings emerge from these results: 1-When it comes to receptive abilities in Catalan and Spanish at the onset of primary school, languages do not take away from each other. That is, higher abilities in one language are related to higher abilities in the other language, resulting in high-performing and low-performing profiles. 2-Despite societal bilingualism in Catalonia, engaging more frequently with the minority language, Catalan, at school and outside may confer advantages not only in Catalan but in Spanish as well.
