Related Experiment Videos
Syntactic gender and semantic expectancy: ERPs reveal early autonomy and late interaction
T C Gunter1, A D Friederici, H Schriefers
1Max-Planck Institute of Cognitive Neuroscience, Leipzig, Germany. gunter@cns.mpg.de
Journal of Cognitive Neuroscience
|August 11, 2000
Summary
This study on German sentence processing found that semantic expectancy influences early language comprehension, while grammatical gender agreement interacts with semantics later. Event-related potentials (ERPs) revealed distinct processing stages.
Area of Science:
- Psycholinguistics
- Cognitive Neuroscience
- Computational Linguistics
Background:
- Understanding how the brain processes language involves examining the interplay between semantic and syntactic information.
- Event-related potentials (ERPs) offer valuable insights into the temporal dynamics of language comprehension.
Purpose of the Study:
- To investigate the independent and interactive effects of semantic expectancy and grammatical gender agreement on language processing in German.
- To determine the timing of semantic and syntactic integration during sentence comprehension.
Main Methods:
- Utilizing event-related potentials (ERPs) to measure brain activity in response to nouns in German sentences.
- Manipulating semantic expectancy (high vs. low cloze probability) and grammatical gender agreement (correct vs. violation).
Main Results:
- Low-cloze nouns elicited a larger N400 component, indicating greater semantic processing load.
- Grammatical gender violations consistently evoked a left-anterior negativity (LAN).
- A P600 component, associated with syntactic reanalysis, was observed only for high-cloze nouns, suggesting an interaction between semantic and syntactic information.
Conclusions:
- Syntactic and semantic processing appear to operate autonomously in early stages of language comprehension.
- Semantic and syntactic information interact during later processing phases, particularly when semantic expectancy is high.
- The findings contribute to models of sentence processing, highlighting distinct temporal dynamics for different linguistic information types.