Second language research using magnetoencephalography: a review

schmidt and roberts

This article first touts MEG as a very useful tool in studying SLA and neurolinguistics, then summarizes previous SLA MEG studies.

MEG is the magnetic version of EEg. Subjects are more comfortable doing an MEG than other neuroimaging devices. But it is expensive and subjects must stay very still.


Phiko and others (2001;2002) found that the L2 is processed more like the L3 than the L1 in early bilinguals.

Zhang and others (2001; 2005) found that Japanese can detect the difference in /r/ and /l/ when there is no vowel afterwards. This suggests "backward masking."


The MEG was used with Japanese and German speakers to determine that both groups recognized kanji "holistically, rather than piecemeal."


Ihara and Kakigi found that Koreans process Hangul and Kana in different parts of their brains.

Phonology can also be studied with the MEG. Differences in pitch changes resulting in changed meaning of words can be taught to SLL. French and Spanish speakers show different areas of the brain allotted to vowels because French has 7 more vowels than Spanish.

Bialystok et al 2005 examined whether bilinguals perform better than monolinguals at the Simon task. The two groups had speedy responses when different areas of the brain were activated. Bilinguals were fast when the "cingulate and superior and inferior frontal regions" and slower with the "right visual cortex." Monolinguals were fast with the "left middle frontal activation" and slow with the "right motor cortex."

Valaki et al (2004) found that Mandarin speakers' brains were less strongly lateralized.