ACTIVATION OF THE HUMAN AUDITORY-CORTEX BY SPEECH SOUNDS

Magnetic evoked responses were recorded to different speech sounds in healthy humans. (i) Short words consisting of fricative consonant/vowel combinations evoked strong responses at the auditory cortex about 100 ms after the vowel onset. The response is specific to acoustic rather than phonetic aspects of the sounds. (ii) In a categorization task, words elicited a transient response followed by a sustained field (SF). When the subject counted the number of target words, SF was clearly increased. There were no consistent differences between the hemispheres and a similar increase of SF was observed when the subject classified the duration of two tones. (iii) When tone 'probes' were presented randomly to either ear and speech sounds to one ear, the 100-ms response was dampened and delayed bilaterally. The dampening was not specific to speech masking but dependent on the amount of frequency and amplitude transitions in the masker. All these experiments suggest that the auditory system performs a very similar analysis of both speech signals and other sounds. (iv) In a recent study, more closely related to speech perception, visual input from articulatory movements of the speaker was found to affect the activity of the auditory cortex. It seems that MEG studies can be useful in the study of brain mechanisms underlying speech perception in intact humans.