Schedule of Events | Symposia

Representational Similarity Analysis of the Neural Codes in Word Reading

Poster Session E - Monday, March 31, 2025, 2:30 – 4:30 pm EDT, Back Bay Ballroom/Republic Ballroom

Deanne Wah¹ (dwah@uwo.ca), Marc Joanisse^1,2; ¹The University of Western Ontario, ²Haskins Laboratories

Contemporary models of word reading suggest that reading occurs using two distinct pathways, a dorsal-route decoding pathway (orthography to phonology to semantics; O-P-S) and a ventral-route direct whole word pathway (orthography to semantics; O-S). Rather than forming distinct processes, the two pathways are suggested to interact via a division of labour, with individual readers perhaps differing in their degree of reliance on one or the other. Classical univariate neuroimaging studies are unable to capture graded representations and interdependence of representational units (such as orthography, phonology, semantics, and the intermediate processes). Thus, the present study aimed to investigate how these neural codes are represented in the reading brain using representational similarity analysis (RSA) and whether individual differences in the strength of representations within pathways predict reading ability. Healthy adult monolingual English readers (N = 50) silently read words during fMRI scanning. An RSA searchlight approach was used to identify different reading sub-processes, as instantiated within a computational model of reading. Individual differences were then identified by correlating RSA results with scores on reading and cognitive tests completed outside of the scanner. We found that O-S processing follows a ventral sight recognition pathway, and semantic representations are distributed across the reading network. Additionally, stronger semantic and O-S representations in ventral stream brain regions were related to individuals’ greater sensitivity to O-S information. The results support the theory that differences in reading skill ensue from the relative engagement of whole-word processes during visual word processing.

Topic Area: LANGUAGE: Other