The Public Library of Science published the study earlier this year. The twenty-one adults in the experiment learned Brocanto2, a thirteen-word language created for the study. The words and grammar rules relate to a computer game similar to chess that the learners played. For example, "Blom neimo lu neep li praz" means "The square blom-piece switches with the neep-piece."

The researchers tested the people three to six months after they had learned the language, to see how well they could remember it. The study found that those who had learned it with the immersion method had brain waves similar to those of native speakers of a language when speaking that language.

Professor Ullman says those who trained with the classroom method also became more native-like in their brain processing. But only the immersion group showed full native-like processing of the grammar. Still, he says teachers should be careful how they use the results of his study.

MICHAEL ULLMAN: "You know I would not make any curriculum changes based on this. Nevertheless, it is suggestive, and I think it warrants further research to see whether in fact what kind of training might in fact be best not just for reaching the native brain bases but also for, you know, maximum proficiency in different aspects of language, like grammar, you know, syntax and lexicon. So I think further research is warranted. And it may be, for example, that a combination of classroom and immersion might be best. But we don't know that."