Do Children Use Multi-Word Information in Real-Time Sentence Comprehension?

Area of Science:

• Psycholinguistics
• Developmental Psychology
• Cognitive Science

Background:

• Children's language acquisition relies on understanding word meaning and frequency.
• While children use single-word cues for predictions, their use of multi-word sequences is less understood.

Purpose of the Study:

• To investigate if children use multi-word sequences for real-time linguistic predictions.
• To determine if children's predictions are based on multi-word information or pairwise associations.
• To explore how children process novel multi-word sequences.

Main Methods:

• Two visual-world eye-tracking experiments with 5- to 8-year-old children.
• Experiment 1 compared predictions for frequent vs. infrequent multi-word sequences.
• Experiment 2 used a training paradigm with novel multi-word sequences.

Main Results:

• Experiment 1 suggested children initially relied on event associations over phrasal frequency.
• Experiment 2 demonstrated children's ability to form predictions from novel multi-word sequences.
• Children's gaze patterns indicated real-time prediction generation.

Conclusions:

• Children utilize multi-word information for linguistic predictions in a context-dependent manner.
• Findings emphasize the importance of multi-word sequences in language processing models.
• This research sheds light on the developmental trajectory of predictive language processing in children.