Orthographic codes in visual word recognition: Task-dependent effects with insertions and deletions of repeated and unique letters

How letter position and identity information in strings is processed has been of great importance for visual word recognition and understanding discrimination between similar words. Position and identity are often not a one-to-one mapping, because in most words at least one letter is repeated, occurring in multiple positions. Whether and how multiple correspondences between identity and position affect reading is not yet clear as repeated letter effects occur inconsistently. Here, we investigated this issue with stimuli constructed from words with repetitions (e.g., REJECTS). We manipulated the presence of different identity information by inserting or deleting repeated (e.g., rejectcs or rjects) or unique letters (e.g., rejectas or rejecs), either as primes for the base words in lexical decision (Experiment 1) or the same-different task (Experiment 3), or as nonword foils in lexical decision (Experiment 2). The deletion of a repeated letter resulted in shorter response times than the deletion of a unique letter, but only in the primed lexical decision task. In contrast, the insertion of a repeated letter resulted in shorter response times than the insertion of a unique letter, but only for the primed same-different and the unprimed lexical decision tasks. Repetition effects were also observed in the accuracy for the nonword foil rejection for both insertions and deletions. These findings provide further evidence for differential processing between repeated and unique letters, while also showing that absent expected repetitions and present redundant repetitions affect processing according to the task, suggesting task idiosyncrasies in the demands, mechanisms or representations involved.