Abstract
Phonetic ambiguities (e.g., a name/an aim) were excised from semantically neutral and disambiguating sentence contexts. The isolates were used to compare the effects of item familiarity, listener expectations, syntactic context, and semantic plausibility on segmentation and identification of the speech stream. In Experiment 1, 36 college students identified 81.6% nonambiguous controls, and 48.9% of the ambiguous isolates. In Experiment 2, 94 students (unaware vs. preinformed of the ambiguities) listened either to isolates or to sentences that were neutral, semantically compatible, or contradictory to the reinserted isolates. Semantic plausibility of sentences shifted perception most powerfully from the acoustic base. Item familiarity interacted with all other variables. Expectations and syntactic effects were intermediate. Results are related to conversational speech perception.
Get full access to this article
View all access options for this article.
