Orthographic and feature-level contributions to letter identification

Abstract

Word recognition is facilitated by primes containing visually similar letters ( dentjst-dentist), suggesting that letter identities are encoded with initial uncertainty. Orthographic knowledge also guides letter identification, as readers are more accurate at identifying letters in words compared with pseudowords. We investigated how high-level orthographic knowledge and low-level visual feature analysis operate in combination during letter identification. We conducted a Reicher–Wheeler task to compare readers’ ability to discriminate between visually similar and dissimilar letters across different orthographic contexts (words, pseudowords, and consonant strings). Orthographic context and visual similarity had independent effects on letter identification, and there was no interaction between these factors. The magnitude of these effects indicated that high-level orthographic information plays a greater role than low-level visual feature information in letter identification. We propose that readers use orthographic knowledge to refine potential letter candidates while visual feature information is accumulated. This combination of high-level knowledge and low-level feature analysis may be essential in permitting the flexibility required to identify visual variations of the same letter (e.g., N-n) while maintaining enough precision to tell visually similar letters apart (e.g., n-h). These results provide new insights on the integration of visual and linguistic information and highlight the need for greater integration between models of reading and visual processing.