Task effects in event related potentials for semantics and gender in French

Gascon, A. 1 , Lebel, V. 1 , Royle, P. 1, 2 , Drury, J. E. . 3 & Steinhauer, K. 4, 2

1 Université de Montréal
2 Centre for Research on Language Mind and Brain
3 Stony Brook, NYU
4 McGill University

We present a study using event-related potentials (ERPs) for gender, and semantic processing in French. We created mismatches between spoken sentences and picture stimuli (1) to elicit ERPs. A number of components were expected in the different conditions.

Picture: White fish in box
1.a “Il y a un CRAYON blanc dans la boîte” ‘there is a white crayon in the box’; Mismatch: Semantic; Expected ERP: N400.
1.b Sentence: “… UNE poisson blanc …” ‘… a.f fish white.m’; Mismatch: Determiner gender; Expected ERPs: AN + P600.
1.c “… un poisson BLANCHE …” ‘a.m fish white.f'; Mismatch: Adjective gender; Expected ERPs: AN + P600.

In preparation of a study with children, we investigated whether these ERPs arise in adults contrasting Task (grammaticality judgment) versus No-Task conditions. We expected task effects to modulate ERP components in different ways. Earlier components (e.g., Anterior Negativities, AN) were not expected to vary, while later, presumably less automatic, signatures (N400, P600) were expected to vary according to Task. Thirteen right-handed adult French-speakers with no history of neurological impairment took part in the study. Six performed the task.

Results: Expected components were elicited in all conditions (see 2). The semantic N400 and the adjective violation AN were not modulated by Task, while the P600s were consistently larger in the Task group for all agreement errors. In addition, the negativity preceding the P600 in the determiner agreement condition varied in scalp distribution. This shows that these designs can be used in studies where no task can be deployed (e.g. children) since expected components arise in the No-Task condition.

2.a Semantic: Task: N400; No-Task: N400; no Task effect.
2.b Determiner gender: Task: Negativity + P600; No-Task: AN + P600; Task effects: Negativity topography + P600 amplitude.
2.c Adjective gender: Task: AN + P600; No-Task: AN + P600; Task effect: P600 amplitude.