Hidden patterns, visible effects? Task structure and cue competition in morphographic spelling

English orthography encodes morphological structure, yet how spellers acquire and use this knowledge remains debated between rule-based and experience-based accounts. We present a three-pronged investigation into the spelling of English pseudowords targeting the morphographic endings ⟨s⟩ and ⟨ous⟩. Experiment 1 (spelling-to-dictation) and Experiment 2 (forced-choice) establish sensitivity to morphographic patterns and test how this sensitivity is modulated by literacy and task structure. We then apply Naive Discriminative Learning to model how morphographic cue weights emerge from statistical regularities in the lexicon and use these weights to reanalyze the behavioral data. Both behavioral studies reveal robust sensitivity to morphosyntactic context, consistently enhanced by literacy and modulated by task structure, with forced-choice responses reducing baseline spelling biases. This pattern suggests that task demands shape how hidden distributional patterns surface in spelling behavior. NDL weights derived from lexical input alone account for variance in participants’ spelling choices, supporting an account in which morphographic spelling reflects probabilistic cue competition driven by productive experience with the writing system.

Manuscript under review