Using global complexity measures to assess second language proficiency: Comparing CLIL and non-CLIL learners of English and Dutch in French-speaking Belgium

This study falls within the framework of an interdisciplinary project on Content and Language Integrated Learning (CLIL) in French-speaking Belgium. One of the project’s aims is to compare the L2 language proficiency of CLIL and non-CLIL French-speaking learners of English and Dutch. In the present paper we focus on learners’ global proficiency and use of different types of metrics to assess syntactic and lexical complexity in the learners’ written productions. Using various computational tools, we extracted lexical and syntactic complexity scores for texts written by CLIL and non-CLIL pupils in their L2 (English or Dutch) and their L1 (French). These scores were then compared to investigate the potential influence of CLIL education on the pupils’ language proficiency as CLIL programs provide more target language input than non-CLIL programs. We therefore hypothesized that CLIL pupils would display a more native-like competence in the target language, i.e. a more native-like level of syntactic and lexical complexity in their writing. As for the influence of CLIL programs on the L1, we did not expect any difference between the two groups. Our results show that our first hypothesis is only partly confirmed as the effect of CLIL on L2 complexity varies according to the language: while the Dutch texts written by CLIL pupils turn out to be more complex for nearly all measures, this was only the case for half of the measures in the English texts. As initially expected for our second hypothesis, we found no influence of CLIL on the complexity of the pupils’ L1