Infants' Developing Sensitivity to Object Function: Attention to Features and Feature Correlations

Abstract

When learning object function, infants must detect relations among features—for example, that squeezing is associated with squeaking or that objects with wheels roll. Previously, Perone and Oakes (2006) found 10-month-old infants were sensitive to relations between object appearances and actions, but not to relations between appearances and sounds or actions and sounds. In this article the authors probed the development of infants' attention to feature correlations critical for representing function by testing 8- and 12-month-old infants' (N = 126) sensitivity to such relations. Eight-month-old infants learned individual features but were not sensitive to the relations between those features. Twelve-month-old infants were sensitive to the relation among the features and significantly responded to violations in learned relations between object appearances and actions and between appearances and sounds. Thus, across development, infants become sensitive to an increasing number of relations with age, supporting an information-processing account of the development of object function.

ACKNOWLEDGMENTS

This work was made possible by National Science Foundation grant BCS 0921634. We thank Sammy Perone for help in various stages of this work, and Shaena Stille, Lisa Christoffer, and undergraduates in the University of California, Davis, Infant Cognition Laboratories, for help with data collection.

Notes

Note. Values in parentheses are standard deviations.

Note. In the appearance/action condition, object appearance and action are correlated across habituation events, and the outcome (sound) is constant across events. In the appearance/sound condition, object appearance and sound are correlated, and the action is constant across events. In the action/sound condition, action and sound are correlated and object appearance is constant across events. The particular features included in the table are for illustration purposes only; in reality, all values of the features were used both during habituation and test, and each infant received different combinations of object appearance, action, and sound.