Low-frequency neural activity tracks syntactic information through semantic mediation.

Area of Science:

• Neuroscience
• Linguistics
• Cognitive Science

Background:

• Understanding how the brain integrates words into linguistic units is crucial in neurolinguistics.
• Prior research primarily focused on semantic or syntactic levels, often failing to separate these factors.
• Limited studies investigated cortical activity tracking word sequences with varying semantic correlations.

Purpose of the Study:

• To investigate how cortical activity tracks word sequences with different semantic correlations.
• To differentiate semantic from syntactic factors in language processing.
• To explore the neural mechanisms underlying word integration in speech perception.

Main Methods:

• Speech perception experiment using electroencephalography (EEG) and the frequency-tagging paradigm.
• Participants (N=25) listened to various word sequences while neural activity was recorded.
• Model simulation using GPT-2 surprisal values to predict neural responses.

Main Results:

• Low-frequency neural activity was found to track syntactic information.
• This tracking of syntactic information occurs through semantic mediation.
• EEG results and model predictions converged, supporting the findings.

Conclusions:

• The brain utilizes low-frequency neural activity for syntactic processing, mediated by semantic context.
• Findings provide insights into the complex language processing mechanism.
• This study advances our understanding of how sequential information is integrated in the brain.