Publications and Research
Document Type
Article
Publication Date
2019
Abstract
In contrast to reports of wild primates, studies of captive primates' flexibility often reveal conservatism: individuals are unable to switch to new and more efficient strategies when task demands change. We propose that such conservatism might be a result of task design and hypothesize that conservatism might be linked to primates' lack of causal understanding in relation to experimental apparatuses. We investigated if chimpanzees (Pan troglodytes) and western lowland gorillas (Gorilla gorilla gorilla) would show greater flexibility when presented with a causally-clear task. We presented six chimpanzees and seven gorillas with a clear tube from which they had to remove straws to release a reward. To first evaluate the apes' causal understanding, we recorded the efficiency with which the apes solved the task (i.e., whether they only removed straws below the reward, ignoring redundant ones above it). To further explore how they solved the task, we also recorded the order in which they removed the straws, which allowed us to determine if habitual action sequences emerged. All apes spontaneously solved the task in their first trial and across repeated trials the majority of their solutions were efficient (median = 90.9%), demonstrating their understanding of the puzzle. There was individual variation in the consistency of straw removal patterns exhibited by the apes, but no ape developed an exclusive habit in the order with which they removed the straws, further indicating their causal understanding of the task. Next, we presented the apes with a new configuration of the same task that required the apes to remove fewer straws to obtain the reward. All apes switched to a more efficient straw removal sequence even though their previously-successful, but now less-efficient, solution remained available. We theorize that because the apes understood the causality of the task, they did not form habits and were not conservative.
Comments
This article was originally published in PeerJ, available at http://doi.org/10.7717/peerj.6195
This work is distributed under a Creative Commons Attribution 4.0 International License (CC BY 4.0).