New FDP paper in the Journal of Comparative Psychology!
Fugazza et al. assessed dogs’ ability to generalize imitation across modifications in context.
Dogs trained to imitate human actions with the Do as I Do method were tested on their ability to reproduce object-related actions demonstrated by their owners. After the demonstration of the action, a retention interval of 1 minute elapsed. During this interval, they displaced the object used by the demonstrator to observe if, when then commanded to imitate (‘Do it!’), the dogs would 1) go to the new location of the object or 2) go to the location where the object was during the demonstration and 3) if dogs would still remember and imitate the action in this new context.
Results showed that dogs went to the correct object and imitated the demonstrated action when only one object was displaced. When the position of two objects was swapped, dogs approached the location where the demonstration was performed and their imitation success dropped significantly.
They repeated this two-objects tests adding pointing gestures towards the correct object or ostensive cues to enhance the attention of the dogs towards the correct object. In these conditions dogs’ imitative performance was restored.
Tthe authors conclude that dogs can generalize imitation to modifications of context to a certain extent. In more difficult conditions dogs’ spatial bias hinders their imitation success. They suggest that this spatial bias is caused by the ostensive communication provided by the demonstrator in the location of demonstration.