Short-term perceptual reweighting in suprasegmental categorization

Area of Science:

• Psycholinguistics
• Auditory Neuroscience
• Speech Perception

Background:

• Speech perception involves integrating multiple acoustic dimensions.
• Perceptual weights of acoustic dimensions dynamically adjust to context.
• It is debated whether suprasegmental speech features are perceived categorically.

Purpose of the Study:

• Investigate dynamic perceptual weighting of acoustic dimensions in suprasegmental speech perception.
• Determine if short-term statistical learning extends to categorical perception of word emphasis.
• Examine the role of fundamental frequency (F0) and duration in processing accented speech.

Main Methods:

• Participants categorized a two-word phrase with typical and atypical covariation of F0 and duration.
• An artificial 'accent' manipulated the correlation between F0 and duration.
• Listeners' categorization decisions were analyzed to infer perceptual weighting.

Main Results:

• Listeners rapidly down-weighted the secondary acoustic dimension (duration) when processing accented speech.
• This demonstrates rapid adaptation to short-term statistical regularities in speech.
• Evidence suggests suprasegmental features are processed categorically.

Conclusions:

• Dimension-based statistical learning is a widespread phenomenon in speech perception.
• Listeners dynamically adjust acoustic evidence weighting for suprasegmental decisions.
• Speech perception extends both segmental and suprasegmental categorization through statistical learning.