Speech fine structure contains critical temporal cues to support speech segmentation.

Area of Science:

• Neuroscience
• Auditory Perception
• Speech Processing

Background:

• Speech segmentation is crucial for speech recognition.
• The speech amplitude envelope (SE) is a primary temporal cue.
• The role of temporal fine structure (TFS) in segmentation is less understood.

Purpose of the Study:

• Investigate the contribution of TFS to speech segmentation.
• Determine if TFS provides temporal information for speech processing.
• Explore the interplay between SE and TFS in auditory perception.

Main Methods:

• Magnetoencephalography (MEG) to measure cortical responses.
• Mutual information analysis to assess temporal information in TFS.
• Behavioral experiments with temporally distorted SE.

Main Results:

• TFS entrains cortical activity in the 3–6 Hz range.
• Temporal information in TFS correlates with SE and spectral change.
• TFS aids speech segmentation and recognition when SE is distorted.
• Spectral resolution is critical for extracting temporal information from TFS.

Conclusions:

• TFS provides significant temporal cues for speech segmentation.
• SE and TFS convey comparable temporal information.
• A synthetic view integrating temporal and spectral domains is necessary for understanding speech segmentation.