After the symposium on multilingual education (see my earlier blog post), I continued email discussions with De Houwer. One of the issues that came up was the concept called “selectivity”: When bilinguals speak, do they (select a language and) activate the language they wish to use or are both languages automatically activated (i.e. “not selected), and they have to suppress the language they do not wish to use? She referred me to “The Mysteries of Bilingualism” by François Grosjean, who listed many studies on this issue to show that the answer is very complex. So many contextual, cognitive, and proficiency factors affect the performance of bilinguals: Sometimes they act as if to take the selective mode, but they act otherwise at other times. In some contexts with monolingual contextual cues, bilinguals’ performance was indistinguishable from that of monolinguals.
I was thinking about all these when I taught an English grammar class to three junior high students the other day. Their English was bad. They had trouble answering simple questions, and their speech was belabored and heavily accented. But I knew they could do better because I had had one of those students in my English conversation class, and she could answer the same kinds of questions without any problem, and her pronunciation was much better. How could this be? I wonder if the contextual cues were at fault. The English conversation class had an English-only policy, and we only spoke English. But in the grammar class, the expected medium of instruction was Japanese, and these students came in expecting to speak in Japanese. In a way, they were primed for Japanese rather than English. Just like the bilinguals, my students might have been primed to speak Japanese, affecting their performance in English.