ABSTRACT
2 investigations of early imitative ability were conducted with cross-sectional samples of infants between 4 and 21 weeks. Study 1 involved 2 rounds of modeling 5 gestures to infants. In Study 2, the number of modeled gestures was reduced to 3, but each one was modeled many more times--for the duration of the infant's attention over a 3-min period. When frequencies of reproduction of a modeled act were compared against baseline frequencies and against average frequencies of the act during periods when other actions were modeled, several main results were obtained. First, the most general finding was that imitative-like matching of modeled gestures was the exception, not the rule, at all ages. Second, even where significantly greater frequencies of a gesture occurred during modeling than during control periods, it was always a partial and incomplete version of the modeled act that was reproduced, not a well-formed copy of the adult's gesture. Finally, where results were consistent with an interpretation of imitation, as with responses to tongue protrusion modeling, the effect was restricted to the youngest ages: 4-6 weeks. Given the restricted evidence for imitation, and the fact that neither linear nor curvilinear growth trends were apparent, the most appropriate explanation for the reproduction of tongue protrusions at the youngest age is to be sought in terms of a reflexive or fixed action pattern type of response.