Early and automatic syntactic processing of person agreement

Abstract

Mismatch negativity, an index of automatic cerebral activity in response to novel stimuli, was used to determine the onset of morphosyntactic processing in French. Stimuli were four two-word sentences made up of a pronoun (first or second person) and a verb (first or second person). Verb forms differed only in the inflectional suffix, which made the sentences either syntactically correct or incorrect. The mismatch negativity response was found to be modulated by the grammaticality of the agreement relation at 50-140 ms after inflection onset, corroborating the previous finding that morphosyntactic processing occurs early and out of the focus of attention. The role of the pronoun-suffix association probabilities in determining the observed timing of morphosyntactic processing is discussed.

Mismatch negativity: A tool for studying morphosyntactic processing?

“…forms. As in the study of Brunellière et al (2007), verbal forms in the present experiment were acoustically identical up to the onset of the final vowel, corresponding to the divergence point. All experimental stimuli in the present experiment had exactly the same duration (609 ms), and the verbal inflections (163 ms) and the pronouns (149 ms) were identical in duration across the present experiment and that of Brunellière et al (2007) (see, Fig.…”

“…The experimental stimuli of the present experiment included the same verb as the study of Brunellière et al (2007), but the two verbal forms of the verb were in the future tense. In the study of Brunellière et al (2007) with a violation of subject-verb agreement in person, the two grammatically correct sentences were nous vivons 'we live' and vous vivez 'you live' and the two incorrect sentences were *nous vivez 'we live' and *vous vivons 'you live'. In all experiments, the grammatical nature of pronoun-verb pairs depended on the final vowel, which we called point of violation (see Fig.…”

“…The recordings were subsequently digitized at a sampling rate of 44.1 kHz (16 bit, mono) and analyzed with Cool Edit 2000 and Praat software programs (http://www.praat.org). Digitized recordings of each word pair were selected so that the recording of all word pairs were best matched in their fundamental frequency (F 0 ), intensity, maximal sound energy and with speech stimuli of Brunellière et al (2007). forms.…”

“…Several MMN studies have investigated morphosyntax by comparing the MMN amplitude after correct sentences to that after sentences with morphosyntactic violations. associated with a violation in pronoun-verb agreement in several languages (Finnish, English, French and German) at 100-200 ms after the violation point Pulvermüller and Shtyrov, 2003;Brunellière et al, 2007;Hasting et al, 2007;Pulvermüller et al, 2008). When investigating case violations between a determiner and its noun in German, Menning and colleagues (2005) also found a MMN effect modulated by the case violation at about 150-200 ms after the onset of the deviant word.…”

“…At present, no MMN study has yet been conducted to test whether morphosyntactic features are represented in morphosyntax. Indeed, all the research, including our earlier research (Brunellière et al, 2007), only investigated the effect of morphosyntactic violations on the MMN wave and did not manipulate the number of morphosyntactic features that were violated.…”