Despite a large number of studies on human language comprehension and its development, the neurocognitive system underlying normal and impaired language processing has yet not been characterized. In particular, it is currently debated whether different subsystems in language comprehension work together in a serial or a parallel fashion. The grant project will use online measures of neuronal activity as revealed by the recording of event-related brain potentials (ERPs), to test a new account to human language processing that emerges from a neurodevelopmental perspective. Specifically, the proposal is guided by the hypothesis that adult language processing is characterized by simultaneously activated neurorepresentational systems at different stages of language comprehension, but that these parallel neuronal networks are acquired sequentially during infancy and childhood. Using ERPs recorded in a series of fragment priming experiments, evidence for the proposed neurodevelopmental model of language comprehension will be obtained in infants, children and adults. Additional analysis techniques, such as the characterization of temporal dynamics of ERPs, will be applied to investigate the assumed functional parallelism in neuronal language processing. The results will serve as a basis for a neurolinguistic model of both language acquisition and the functional organization of the adult language processing system. The work will be of importance for questions of education and for the understanding and treatment of developmental language disorders.