Speech register influences listeners' word expectations.

Area of Science:

• Cognitive Neuroscience
• Psycholinguistics
• Computational Linguistics

Background:

• Predictive language processing is crucial for fluent comprehension.
• The influence of broader contextual information, like speech register, on prediction remains under-investigated.
• Previous research often focuses on shorter, controlled linguistic contexts.

Purpose of the Study:

• To investigate how different speech registers (dialogues, news, books) affect predictive language processing.
• To determine if listeners' predictions are driven by general language statistics, register-specific patterns, or recent word occurrences.
• To examine the neural correlates (N400 effect) of register adaptation in real-time speech comprehension.

Main Methods:

• Recorded electroencephalography (EEG) signals from participants listening to naturalistic speech.
• Utilized a statistical language model to calculate word surprisal reflecting different predictive strategies (generic, register-specific, recency-based).
• Analyzed the N400 event-related potential amplitude in relation to word surprisal estimates.

Main Results:

• The N400 amplitude was most strongly predicted by register-specific word surprisal.
• This indicates that the broader linguistic context (speech register) significantly influences predictive language processing.
• Findings suggest that adaptation to speech register goes beyond simple recency effects.

Conclusions:

• Listeners actively adapt their word anticipation strategies to the specific speech register.
• Predictive language processing is sensitive to the statistical properties of the wider linguistic environment.
• This research highlights the dynamic and context-dependent nature of human language comprehension.