P. Helenius, A. Hultén, K. Hytönen, A. Jalava, J. Kujala, H. Laaksonen, M. Liljeström, T. Parviainen, J. Polkko, A. Puurula, T. Saarinen, R. Salmelin, J. Uusvuori, M. Vihla, and M. Viinikainen
MEG is an invaluable tool in characterizing the successive and largely overlapping stages in language processing, from sensory analysis (visual, auditory, tactile) to linguistic assessment, memory search, and motor function.
We have previously shown that when fluently speaking individuals read words out loud activation proceeds from Broca’s area to motor cortex. In developmental stutterers, this order is reversed; abnormalities seem to be limited to processes specifically involved in overt speech production rather than core linguistic analysis. Recently, we found that these ‘production’ areas also show activation while subjects are only listening to spoken sentences, but know that they will need to repeat or transform them after a short delay. In stutterers, abnormalities again emerged in the same areas as in the production task. Task-related modulation of the motor cortical 20-Hz rhythm has further demonstrated strong emphasis of face area in fluent speech production but additional involvement of the adjacent hand area in stutterers. Our recent data on fluent speakers demonstrated involvement of the hand area specifically in production of non-speech mouth movements, regardless of sequence complexity. These data suggest that motor cortical specialization for verbal mouth movements may not have developed normally in individuals who stutter.
In speech perception, lexical-semantic processing is reflected in sustained, bilateral activation of the superior temporal cortex from ~200 ms onwards. We have recently shown that both meaningful linguistic context (top-down) and acoustic-phonetic (bottom-up) cues enhance speech-sensitive analysis at the onset of this sustained response, at 200-300 ms. In dyslexic individuals, onset of the lexical-semantic processing stage is delayed by about 50 ms. It would be tempting to interpret the prelexical abnormality as cortical reflection of impaired phonological processing in dyslexia. However, it could equally well reflect irregularities in basic auditory processing. Interestingly, our recent comparison of speech vs. nonspeech analysis in dyslexia points to abnormalities in general auditory processing that occur in the time window (~100 ms) when phonetic information is extracted. Such an account could help to reconcile the existing, seemingly disparate reports of either auditory or phonological impairments in dyslexia.