drinking and driving

[orbacht]

This is a very well designed study. For once I thought that all three of the experiments were necessary and told us something different! I thought that imitation was a great way to get at infant’s understanding of concepts in a non-verbal way. Imitation really shows that the kids have an internalized understanding of these concepts.
I really liked that the authors managed to prove that infants divide things they see into very broad categories such as animals and vehicles. This is inline with the previous reading we have done that discussed how infants pick up on intentional cues such as self propelled motion and eye gaze. It also reminded me of the glove study where infants recognized a naked hand as belonging to a human but not a hand wearing a metallic glove. Clearly infants are picking up on some broad cues that are allowing them to distinguish between machines and humans. This seems like a far more adaptive strategy than relying on base level cues.
I also found it very interesting that infants did not choose the correct action more often for exemplars that were most similar to modeled exemplar in experiment 1 and 2. This is exactly the opposite of what I would have expected according to prototype theory. I would have thought that children, who have limited experience, would have been more successful at selecting the correct action for the less extreme exemplars. The fact that they got it right regardless of how similar the exemplars were to the model shows that infants are making a judgment about the qualities of the objects on the spot, not referring to previous knowledge.
I have to wonder whether the types of actions the vehicles and animals were made to do are comparable. The animals were acting with agency and intent, drinking and sleeping, whereas the vehicles were being acted upon, being ridden or opened with a key. This doesn’t necessarily mean that these two types of things wouldn’t be in separate categories. In fact this could be a distinguishing factor.
The third experiment was my favourite experiment. The fact that infants choose to perform the action with the “correct” object even when both objects are modeled performing the action showed that infants are not relying on any information given by the demonstrator to draw their conclusions. This could be generalized to other contexts. Infants may not be relying on imitation and experience to draw conclusions about categorization in their everyday lives. This would be a compelling argument for an innate understanding of at least some variables that would help in categorizing objects.