Neural networks underlying contributions from semantics in reading aloud.

Area of Science:

• Neuroscience
• Cognitive Science
• Psycholinguistics

Background:

• Reading involves complex cognitive processes, but the neural underpinnings are not fully understood.
• Existing models like connectionist and dual-route cascaded models offer differing explanations for reading, particularly regarding the role of semantics.
• Evidence supporting these models remains inconclusive, necessitating advanced analytical approaches.

Purpose of the Study:

• To investigate the network-level interactions between semantic and phonological processing regions during reading.
• To differentiate between connectionist and dual-route models using effective connectivity analysis.
• To explore how word properties influence these neural interactions and reading performance.

Main Methods:

• Utilized a novel effective connectivity algorithm (IMaGES) on functional magnetic resonance imaging (fMRI) data.
• Defined regions of interest in the left hemisphere, including semantic areas (angular gyrus, inferior temporal sulcus) and phonological areas (posterior superior temporal gyrus, posterior middle temporal gyrus).
• Analyzed brain activity while participants read aloud words varying in spelling-sound consistency, frequency, and imageability.

Main Results:

• The connectionist model's prediction of increased semantic contribution for low-consistency words was supported.
• Effective connectivity analyses showed dynamic interactions between semantic (e.g., ITS) and phonological (e.g., pSTG) regions, modulated by word properties.
• Connectivity from semantic to phonological areas was enhanced for high-imageability and low-consistency words, and individual differences in connectivity correlated with reading performance.

Conclusions:

• The findings support a connectionist view of reading, emphasizing the role of semantic processing.
• Neural pathways between semantic and phonological areas are dynamically engaged during reading aloud.
• The strength of semantic-phonological connectivity is associated with optimal reading performance, highlighting its importance in the reading network.