skip to primary navigation skip to content

CBSU bibliography search

To request a reprint of a CBSU publication, please click here to send us an email (reprints may not be available for all publications)

Automatic grammar processing in the brain
SHTYROV, Y., PULVERMULLER, F., Naatanen, R. & Ilmoniemi, R.J.
ICON8, 8th International Conference on Cognitive Neuroscience, September 9-15, 2002, Porquerolles Island, France, p 65.
Year of publication:
CBU number:
We investigated the magnetic correlate of the Mismatch Negativity (MMNm) to syntactically correct and incorrect word strings. The MMNm is a tool for probing neuronal memory traces underlying the processing of speech and language [1-4]. To overcome problems of earlier studies of grammar processing in the brain, we monitored brain activity related to syntactic processing (a) in the absence of directed attention towards the stimuli, (b) without any linguistic task that may invite subjects to apply different linguistic strategies in the processing of grammatical and ungrammatical strings, (c) under strict control for physical and lexical stimulus parameters, (d) using highly sensitive state-of-the art neuroimaging and source localisation techniques. Using a whole-head 306-channel MEG system, we recorded the MMNm elicited either by grammatically correct or ungrammatical phrases, while the subjects were asked to ignore the auditory stimulation and engage in a different activity. As stimuli, we used short phrases differing only in one phoneme, which rendered them as either grammatical or ungrammatical. Any possible effects of acoustic (phonetic) or lexical differences were ruled out by using an orthogonal design reversing the effect of the phoneme combination on grammaticality. The data were analysed using single-dipole models and minimum current estimates (MCEs). We found that syntactically incorrect deviant stimuli elicited larger MMNm responses than the grammatical phrases. This difference was not affected by between-word differences, phonetic contrasts, or early acoustic cues. MCEs indicated that grammatical and ungrammatical strings differentially activated the left superior-temporal cortex, and possibly additional inferior frontal areas. Our results are compatible with earlier findings [5,6]. The MMNm modulation by grammaticality in the absence of directed attention to the stimuli suggests that syntax processing in the brain can be automatic. [1] Natanen, R. et al. Nature 385, 432-434 (1997). [2] Pulvermuller, F., Kujala, T., Shtyrov, Y. et al. NeuroImage 14, 607-616 (2001). [3] Shtyrov, Y. et al. NeuroImage 12, 657-663 (2000). [4] Shtyrov, Y. & Pulvermuller, F. European Journal of Neuroscience (2002, in press). [5] Neville, H. et al. Journal of Cognitive Neuroscience 3, 151-165 (1991). [6] Osterhout, L. et al. Trends in Cognitive Sciences 1,203-209 (1997).