A reexamination of object permanence in domestic dogs

Abstract

Object permanence is the primary cognitive milestone achieved by non-verbal human infants and has been an obvious focus for comparative studies. There are inherent challenges to adapting the Piagetian framework to other species. Previous studies on domestic dogs have failed to control for the effects of olfaction and learning which could confound the interpretation of results. My experimental design controls for both of these effects and reexamines object permanence in domestic dogs. I found that domestic dogs as a species are capable of achieving a fully developed object concept (36 %), although I observed great diversity in performance. I found that the effects of breed, sex, and age differences could not account for this diversity. In contrast to previous studies, I found that dogs as young as four months showed complete object permanence, in contrast to previously published findings. Through random order task presentation in single trials, I rigorously examined ordinality, and found that it was highly conserved in 98.8% of trials. Additionally, this method shows that object permanence cannot be attributed to learning or previous experience. I found further evidence supporting ordinality of stage acquisition by examining the mistakes dogs made a higher order tasks. I observed the A not B error in two dogs at the appropriate stage, in contrast with previous studies. I also found that dogs at stage 5b showed stage 5 search behavior, and dogs at stage 6a showed stage 6 search behavior (p < 0.001) which demonstrates the theorized operational schema of particular cognitive stages. This has not been reported previously in dogs. I found that there was no difference in search times between visible and invisible displacements, but unsuccessful trials were significantly longer than successful trials. Since I only included active search times in this analysis, this difference alludes to metacognition, i.e., dogs' knew that they did not know how to solve the tasks. Search behaviors were also different between trials with different outcomes, which further supports the metacognition hypothesis. Overall, my experiment was the first to effectively address biological concerns with the application of cognitive tasks to domestic dogs.