Abstract
Phonetic ambiguities, such as ice cream/I scream, have been described as perceptually indeterminate. These ambiguous forms have been used to illustrate the requirement of multiple levels of processing (analysis of surrounding syntax and semantics) in the determination of word perception. In examining that requirement, we found both in perception (31 listeners) and in generation (129 college students) that phonetically ambiguous pairs were not confusingly reversible. Listeners of clearly presented tokens were unable to identify pair members (89% failure) except when the alternatives were made explicit (67% and 69% success). Readers took over a minute to locate and identify the alternative segmentations to presented tokens, and language-context effects were negligible. The results support unitary rather than exhaustive processing, and the concepts of word frequency and listener expectations provided a more comprehensive interpretation than did top-down multiple levels of analysis.
Get full access to this article
View all access options for this article.
