Open Access

Reading is an essential ability to master everyday life in our society. The ability to read is based on specific connections between brain regions involved in the reading process – so-called cortical networks for reading. These cortical networks for reading allow us to learn the correct identification of visual words. The use of visual words is based on knowledge about the orthography (lexical) and the meaning of words (semantic). This knowledge must be acquired by beginning readers (first grader), i.e. beginning readers learn in a first step to link letters to a whole word and in a second step associate this whole word with meaning. To retrieve this knowledge during visual word recognition (VWR) a cortical network for lexical-semantic process must be activated. However, it is currently unclear whether beginning readers and reading experts activate the same neuronal network during VWR. Therefore, the aim of this thesis was to investigate the question whether beginning readers (first grader, children) and reading experts (adults) use different cortical networks for the lexical-semantic processing in VWR. To address this question we recorded electroencephalographic (EEG) activity during VWR in children and adults. Children and adults were instructed to read a visualizable word to compare this word with a following picture stimulus. The first part of this thesis is concerned with the analysis of ERPs for visual word recognition in children and adults at sensor level. For both groups we observed the typical ERP components P100 and N170 for visual word recognition. These components differed in amplitude and time course between both groups. The second part of this thesis investigated the neuronal generators (brain areas) of ERPs during VWR and possible differences between children and adults at source level. We observed a high overlap in brain areas involved during VWR in children and adults. However, the brain areas differed in activation and time course between children and adults. Finally, the third and most important part of the thesis investigated the question whether children and adults use different cortical networks for the lexical-semantic processing in VWR over time. To address this question Dynamic Causal Modeling (DCM) and Bayesian model comparison were used. We compared nine biologically plausible cortical network models underlying the ventral lexical-semantic path in VWR. In addition, increasing time intervals were used to consider possible changes of network structure during VWR. The network models included eight brain regions (four bilateral pairs) involved in the lexical-semantic processing in VWR: occipital cortex (OC), temporo-occipital part of inferior temporal gyrus (ITG), temporal pole (TP), and inferior frontal gyrus (IFG). In almost all time intervals we found evidence that children and adults use the same cortical networks for the lexical-semantic processing in VWR. However, we found differences between adults and children in the connection strengths of the favoured model. Interestingly, we found a stronger direct connection from OC to IFG in adults compared to children. In conclusion, our results suggest that children and adults activate largely the same lexical-semantic networks during VWR over time. This supports the notion that children and adults use the same biological fiber connections for VWR. However in contrast to children, adults showed increased use of the shortcut pathway from OC to IFG. The increased use of the shortcut pathway from OC to IFG in adults can be interpreted as consequence of learning. Learning causes in accordance with the Hebbian learning rule (“neurons that fire together, wire together” (Hebb, 1949)) synaptic change. Consequently the frequent coactivation of the input and output stage of OC and IFG during the lexical-semantic process facilitates the stronger direct connection between both brain areas. The stronger direct connection from OC to IFG most likely allows adult reading experts to speed up the lexical-semantic process during VWR. Accordingly, we conclude that the stronger direct connections from OC to IFG in adults compared to children underlay the different reading capabilities in both groups.