Task and Timing Effects in Argument Role Sensitivity: Evidence From Production, EEG, and Computational Modeling.

Area of Science:

• Psycholinguistics
• Cognitive Neuroscience
• Computational Linguistics

Background:

• Readers form expectations about upcoming words using contextual cues.
• Previous research indicated argument roles don't affect neural or behavioral prediction measures.
• Some studies suggested argument roles matter in production or comprehension with extra time.

Purpose of the Study:

• To investigate how task type and prediction time influence lexical prediction sensitivity to argument roles.
• To differentiate between effects of task demands versus temporal constraints on predictive language processing.

Main Methods:

• Used electroencephalogram (EEG) and speeded cloze experiments with matched Japanese stimuli.
• Manipulated prediction time in EEG by altering stimulus intervals.
• Conducted a speeded cloze task matching EEG stimuli and timing.

Main Results:

• EEG with extra prediction time and speeded cloze showed sensitivity to argument roles.
• EEG with limited prediction time replicated argument role insensitivity.
• Both task and timing independently influenced lexical prediction.

Conclusions:

• Lexical prediction is initially argument role-insensitive.
• A serial monitoring mechanism inhibits inappropriate candidates, operating faster in production than comprehension.
• This mechanism explains observed task and timing effects in predictive language processing.