Cognitive processing across verbal and pictorial domains: ERP evidence from adults, children with typical language development, and children with developmental language disorder

Abstract in English

Linguistic theories propose that cognitive processing of language is either specific to the verbal domain or domain general. These theories lead to contrasting predictions, which are explored in this dissertation. The aim is to advance the understanding of children’s cognitive processing in the verbal domain in comparison to that in the pictorial domain. Previous findings indicate shared features between cognitive processing in these domains, though mainly in adults and in separate studies. In contrast to earlier studies, the papers on which this dissertation is based have a within-subjects design that isolates the impact of domain (verbal/pictorial) and include both children with typical language development (TD) and peers with developmental language disorder (DLD). Cognitive processing was assessed by measuring event-related potentials (ERPs), which are averaged electrical brain responses obtained by electroencephalography (EEG). In Paper I and II, we manipulated semantic predictability in both domains, that is, of words and pictures in their respective narrative context. Children with TD showed the expected N400 ERP effect in both domains, indicating semantic predictions. However, the N400 effect was not discernible in children with DLD in either domain, suggesting challenges in making semantic predictions across verbal and pictorial domains. In Paper III, we manipulated the hierarchical structure of both sentences and picture sequences, by inserting prolonged silences and blank panels, respectively, both within and between constituents. To validate the novel manipulation in the verbal domain, Paper III included adults, whose ERP responses aligned with prior findings in the pictorial domain. Across both domains, children with TD did not show the expected increased anterior negativity for within constituent disruptions. This could indicate that children may not activate long-term memory representations of hierarchical structure in an adult manner irrespective of domain. Together, Papers I–III reveal multiple similarities in cognitive processing across the verbal and pictorial domains. These findings offer novel support for domaingeneral perspectives on language processing. A key implication is that picture sequences may place similar demands on comprehension as sentences, which could be especially relevant to consider in communication with children with DLD.