Semantic Context Enhances the Early Auditory Encoding of Natural Speech

Area of Science:

• Neuroscience
• Cognitive Science
• Psycholinguistics

Background:

• Speech perception integrates sensory input with contextual expectations.
• The role of prior knowledge in early auditory encoding versus feedforward processing is debated.
• Limited research exists using naturalistic speech stimuli to investigate this debate.

Purpose of the Study:

• To investigate how semantic context influences low-level auditory encoding of natural speech.
• To determine if top-down prior knowledge affects early cortical processing of speech.
• To explore the relationship between semantic context quantification and auditory encoding indexing.

Main Methods:

• Recorded electroencephalography (EEG) from participants listening to continuous natural speech.
• Quantified semantic context of speech using a novel method.
• Indexed low-level auditory encoding using a common method.
• Employed a forward modeling approach to assess prediction accuracy of EEG signals.

Main Results:

• Early cortical tracking of the speech envelope was enhanced by semantic similarity to sentential context.
• Prediction accuracy of EEG signals mirrored this enhancement effect.
• Temporal patterns of correlation varied based on acoustic or phonological representations, suggesting top-down information flow.

Conclusions:

• Semantic context enhances the early cortical tracking of words in natural speech.
• Findings support a mechanism linking top-down prior information with early cortical entrainment.
• This suggests that prior knowledge influences bottom-up sensory processing during speech comprehension.