Use this URL to cite or link to this record in EThOS:
Title: Early visual contributions to reading
Author: Hogan, John S.
ISNI:       0000 0004 7659 9654
Awarding Body: UCL (University College London)
Current Institution: University College London (University of London)
Date of Award: 2018
Availability of Full Text:
Access from EThOS:
Full text unavailable from EThOS. Please try the link below.
Access from Institution:
Reading requires integrating visual and linguistic processes, so it is perhaps surprising that models of visual word recognition focus almost entirely on language, to the exclusion of vision. Neurological models of reading assume that visual information proceeds serially from the retina through the early visual cortices, where a hierarchy of increasingly complex feature detectors transform the sensory-bound retinotopic code into progressively more abstract forms, eventually reaching reading-specialised populations that encode orthographic units. These orthographic detectors are the input pathway to the wider language system. This notion of a serial staged hierarchy culminating in abstract detectors, however, is likely overly simplistic accumulating. Evidence suggests that the occipitotemporal system is better understood as a highly recurrent network. Classical hierarchical accounts and interactive processing models make contrasting predictions about how early visual areas contribute to reading. To test these predictions, I retinotopically mapped occipital visual areas and then measured (fMRI) their neural response to different reading tasks. I found that reading strongly engaged areas V1-V3 bilaterally, both in the central (stimulated) regions and in regions coding the periphery, suggesting both bottom-up and top-down influences in early visual cortices. Within the central regions of V1-V3, activity was significantly stronger for low frequency than high frequency words, again suggesting that non-visual factors such as lexical frequency influence processing in the earliest visual areas. Subsequent analyses revealed that ventral (V4, VO-1, VO-2) and dorsal (V3a, V3b, V7) regions were both active during reading, with no evidence of any difference in the strength of activation between them. Lastly, I found that a seemingly incidental property of the experimental paradigm (stimulus presentation rate) dramatically affected V1-V3 activity. Together these results contradict the notion that the early visual stages of reading are tightly sensory-bound and require reading-specific neuronal representations. Rather, these findings suggest that reading is an interactive process, even in the earliest visual cortices.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available