Date created: | Last Updated:

Category: Project

Description: Speech unfolds rapidly, and information necessary for phoneme recognition does not arrive simultaneously. Listeners process speech incrementally: information is used as soon as it arrives to make partial decisions and anticipate future material. However, fricatives are not processed incrementally: listeners delay lexical access until all cues arrive. Fricatives also contain coarticulation that anticipates upcoming vowels. We examined the timecourse of anticipation, finding that listeners anticipate vowels immediately, but wait several hundred milliseconds for fricative identification. Thus, listeners do not process speech in the order that cues arrive—the dynamics of language processing may be only loosely coupled to the input.