An ERP study of structural anomalies in native and semantic free artificial grammar: Evidence for shared processing mechanisms

• Subjects were trained in a semantic-free artificial grammar.
• We studied ERPs elicited by the artificial grammar's correct vs. incorrect sequences.
• We studied ERPs elicited by correct vs. incorrect sentences in native language.
• ERPs of both grammars showed very similar posterior negativities and P600 effects.
• P600 might reflect costs of structural prediction/integration in ordered sequences.

Artificial grammars have been widely applied to the study of sequential learning in language, but few studies have directly compared the neural correlates of artificial and native grammar processing. In this study, we examined Event Related Potentials (ERPs) elicited by structural anomalies in semantic-free artificial grammar sequences and sentences in the subjects’ native language (Spanish). Although ERPs differed during early stages, we observed similar posterior negativities (N400) and P600 effects in a late stage. We interpret these results as evidence of at least partially shared neural mechanisms for processing of language and artificial grammars. We suggest that in both the natural and artificial grammars, the N400 and P600 components we observed can be explained as the result of unfulfilled predictions about incoming stimuli.