A hierarchy of linguistic predictions during natural language comprehension
Abstract
Theorists propose that the brain constantly generates implicit predictions that guide information processing. During language comprehension, such predictions have indeed been observed, but it remains disputed under which conditions and at which processing level these predictions occur. Here, we address both questions by analyzing brain recordings of participants listening to audiobooks, and using a deep neural network to quantify the predictions evoked by the story. We find that brain responses are continuously modulated by linguistic predictions. We observe predictions at the level of meaning, grammar, words, and speech sounds, and find that high-level predictions can inform low-level ones. These results establish the predictive nature of language processing, demonstrating that the brain spontaneously predicts upcoming language at multiple levels of abstraction.
- Publication:
-
Proceedings of the National Academy of Science
- Pub Date:
- August 2022
- DOI:
- 10.1073/pnas.2201968119
- Bibcode:
- 2022PNAS..11901968H