“Assembled” Phonology and Reading: A Case Study in How Theoretical Perspective Shapes Empirical Investigation

This chapter report experiments that used three laboratory reading tasks such as new lexical decision, semantic categorization, and proofreading experiments. All produced large reliable effects of nonword phonology. If performance in laboratory tasks pertains to typical meaningful experience of text, then it should not be peculiar to a single task. Subjects in a lexical decision experiment judge whether individually presented letter strings are words. They respond “word” to letter strings that are words and “nonword” otherwise. The lexical decision task is used to test whether nonword stimuli falsely retrieve lexical memories of words similar in orthography and phonology. The method detects whether correct rejection of a nonword foil like SLEAT included retrieval of lexical memory for SLEET. Subjects in the lexical decision task judged whether individually presented letter strings were words. Ideally, the lexical decision task isolates word recognition because of the isolated presentation of letter strings, and because words merely need to be recognized for correct performance. But not surprisingly, perhaps, the lexical decision task falls short of this ideal. It requires that words should be discriminated from nonwords, not just recognized. Thus, effects thought to originate in recognition and retrieval is confounded with effects of (possibly) task-specific processes of discrimination.