Testing Bilingual Educational Methods: A Plea to End the Language-Mixing Taboo

Abstract

Language mixing in a given class is often avoided in bilingual education because of the generally held belief that one subject should be taught in only one language and one person should stick to one language in order to minimize confusion. Here, we compared the effects of mixing two languages and monolingual functioning on memory performance in immediate recall as a proxy for comprehension and attention during learning. In Experiment 1, nonbalanced bilingual youngsters were provided with definition pairs introducing familiar objects in a single-language context (SLC) or in a mixed-language context (MLC). After each definition block, participants were asked to identify previously introduced objects presented among a stream of Old and New items. In Experiment 2, the same speaker produced the two definitions in the mixed-language context, thus violating the second principle introduced above. In both experiments we found no advantage for the SLC over the MLC of exposure.