Social Inference May Guide Early Lexical Learning.

Area of Science:

• Developmental Psychology
• Computational Linguistics
• Cognitive Science

Background:

• Infant word learning is influenced by social cues.
• Previous models often overlook the role of social categorization in language acquisition.

Purpose of the Study:

• To incorporate social reasoning about informant groups into a computational model of word learning.
• To test if this model can account for infant looking behavior in word learning and recognition tasks.
• To explore the interaction between social and linguistic development.

Main Methods:

• Developed a computational model integrating social reasoning with word learning.
• Simulated infant behavior in two experimental paradigms: object labeling and the Switch task.
• Analyzed how infant age and talker variability affect performance on distinguishing minimal pair labels.

Main Results:

• The model successfully accounts for infant looking behavior in both word learning and recognition tasks.
• Simulation 1 showed the model's ability to predict infant responses to correct/incorrect labels from different talker types.
• Simulation 2 explained how multiple talkers improve 14-month-olds' performance on the Switch task, especially with minimal pair discrimination.

Conclusions:

• Infant looking behavior in word learning tasks is guided by beliefs about talker knowledgeability and group membership.
• Social categorization significantly impacts linguistic development in realistic, diverse communication settings.
• Findings underscore the intricate interplay between social cognition and language acquisition in early development.