Predictions about prosody facilitate lexical access: Evidence from P50/N100 and MMN components

Area of Science:

• Cognitive Neuroscience
• Psycholinguistics
• Auditory Perception

Background:

• Predictive coding models suggest top-down predictions influence sensory processing.
• The exact mechanisms of predictive coding in speech perception, especially concerning prosody and lexical access, are still debated.
• Simultaneous predictions in different linguistic domains (e.g., prosodic, semantic) require further investigation.

Purpose of the Study:

• To investigate how predictions about prosody facilitate speech perception.
• To examine lexical access influenced by simultaneous prosodic and semantic predictions.
• To contribute to the understanding of the predictive coding framework in auditory processing.

Main Methods:

• Utilized a passive auditory oddball paradigm with Dutch participants.
• Examined neural responses (P50/N100, MMN) to prosodic changes in real words and acoustically matched pseudowords.
• Compared responses to prosodic changes that also induced semantic changes in real words.

Main Results:

• An enhanced P50/N100 complex was observed for prosodic changes in pseudowords.
• A mismatch negativity (MMN) response was detected for both words and pseudowords.
• A reduced P50/N100 response to prosodic changes in real words was noted, potentially indicating information suppression.

Conclusions:

• The findings align with a multi-level predictive coding framework in speech perception.
• Early neural processing (as early as 50 ms) utilizes predictions about linguistic prosodic information.
• The brain integrates prosodic predictions with sensory and semantic input, as evidenced by MMN responses.