Temporal integration of segmented-speech sounds probed with mismatch negativity

Area of Science:

• Auditory Neuroscience
• Cognitive Psychology
• Event-Related Potentials

Background:

• Mismatch negativity (MMN) is an event-related potential elicited by deviant sounds in a sequence of standard sounds, occurring automatically without attention.
• Auditory sensory memory encodes standard sounds within a temporal window of integration (160-170 ms), a concept established for pure tones but less so for speech.
• Pure tones are primarily processed in the right hemisphere, while speech sounds are predominantly recognized in the left hemisphere.

Purpose of the Study:

• To investigate whether segmented-speech sounds are processed as a temporal unit, analogous to pure tone sounds.
• To explore potential differences in the processing of speech sounds between the right and left auditory pathways.

Main Methods:

• Twenty-five right-handed healthy Japanese men participated.
• Stimuli comprised a Japanese female /a/ vowel, presented in randomized sequences of standard and deviant durations (short to long).
• Stimuli were delivered binaurally (to both ears) separately, and mismatch negativity (MMN) was recorded.

Main Results:

• Definite MMN was observed for all bilateral stimuli, with comparable peak latencies.
• MMN amplitude increased progressively with deviant sound duration (short to long).
• No significant differences in MMN were found between stimuli presented to the right versus the left ear.

Conclusions:

• Bilateral deviant segmented-speech sounds appear to be processed as a temporal unit.
• Auditory processing of segmented speech does not show hemispheric asymmetry in this context.
• The findings support the concept of a consistent temporal integration window for speech sounds across both auditory pathways.