Abstract
One hundred and twenty Caucasian, typically developing 12- and 18-month-old infants were presented with four models that demonstrated novel actions on objects. The model conditions were: (1) a wholly visible person; (2) disembodied human hands; (3) disembodied hands wearing mittens; and (4) disembodied mechanical pincers. Eight target actions on different objects were demonstrated in each condition. A baseline control condition was also included, in which objects were presented but no actions were demonstrated. The 12-month-olds only copied target actions above baseline when demonstrated by a person or by disembodied hands. The 18-month-olds copied all models above baseline; however, at the same time they copied the person and hands significantly more often than the other two models.
Acknowledgments
We thank Mark Nielsen for providing valuable feedback on this manuscript.