Letter identity and position coding in the parafovea.

Area of Science:

• Cognitive Psychology
• Computational Neuroscience
• Visual Perception

Background:

• Letter position coding is crucial for word identification.
• Existing computational models primarily explain foveal word recognition.
• Parafoveal processing, essential for reading, remains less understood.

Purpose of the Study:

• To investigate the flexibility of parafoveal letter identity and position coding.
• To test how retinal eccentricity and visual field influence parafoveal processing.
• To evaluate current reading models against new experimental data.

Main Methods:

• Utilized a same-different match task with transposed- and substituted-letter strings.
• Manipulated stimulus presentation time (100 ms vs. 300 ms) to control for eye movements.
• Varied retinal eccentricities and visual fields (left vs. right) for word stimuli.

Main Results:

• Parafoveal letter positions are harder to code than identities.
• A right visual field benefit was observed in word identification.
• Letter position uncertainty interacted with eccentricity and visual field, with specific patterns for initial vs. final transpositions.

Conclusions:

• Parafoveal processing exhibits greater flexibility than current models suggest.
• Existing reading models struggle to account for the complex interactions observed.
• Findings necessitate revisions to computational models of reading and visual word recognition.