Related Experiment Video
Updated: Jun 5, 2026
05:38
Interaction between Phonological and Semantic Processes in Visual Word Recognition using Electrophysiology
Published on: June 29, 2021
Top-down modulation of ventral occipito-temporal responses during visual word recognition
Tae Twomey1, Keith J Kawabata Duncan, Cathy J Price
1Cognitive, Perceptual & Brain Sciences, University College London, Gower Street, London, WC1E 6BT, UK. t.twomey@ucl.ac.uk
Neuroimage
|January 15, 2011
Summary
Neural models of reading often overlook interactivity. This study shows the left ventral occipito-temporal cortex (vOT) integrates visual and nonvisual information during word recognition, supporting interactive reading models.
Area of Science:
- Neuroscience
- Cognitive Science
- Computational Linguistics
Background:
- Cognitive models emphasize reading interactivity, but neural models often focus on feed-forward processing.
- The left ventral occipito-temporal (vOT) cortex is traditionally viewed as an early visual word recognition stage.
- Limited research explores top-down influences on vOT during reading.
Purpose of the Study:
- To investigate top-down influences on vOT activation during visual word recognition.
- To determine if vOT integrates visual and nonvisual stimulus properties.
- To test the hypothesis of vOT as a visual-semantic/phonological interface.
Main Methods:
- Functional magnetic resonance imaging (fMRI) was employed.
- Participants performed lexical decision tasks focusing on visual or nonvisual properties.
- Stimulus and task variations were used to probe vOT modulation.
Main Results:
- vOT activation showed significant modulation by both stimulus and task manipulations.
- These modulations indicate top-down processing of nonvisual information.
- Evidence supports interactive processing within vOT.
Conclusions:
- The vOT cortex functions as an interface, connecting visual word forms with nonvisual information (semantics, phonology).
- Findings support interactive processing in reading, aligning with cognitive models.
- Results challenge serial, feed-forward assumptions in current neuro-anatomical reading models.
