Cognitive flexibility in gibbons (Hylobatidae): object manipulation and tool-use

Abstract

Gibbons (Hylobatidae), taxonomically apes, have been largely ignored in cognitive research. This is surprising given their unique phylogenetic position, being intermediate between the monkeys and great apes, and the available diversity of extant species. They are therefore, ideally placed to study the evolution of cognitive abilities in the hominoid line; they offer the opportunity to determine how the mental capacities of primates have changed through the transition from monkey to ape. This research aimed to begin to fill the void in our knowledge regarding the cognitive abilities of this family through investigations of their object manipulation and tool-use skills, relating the findings to the evolution of the hominoid brain.

In a raking-in task, where the gibbons were presented with a tool that could be used to draw in an out-of-reach food item, these apes evidenced potentially insightful comprehension of object relationships when the tool and goal object were presented in direct alignment. They also proficiently used a rake to retrieve a reward while avoiding a trap that presented an impediment to goal attainment; however, in general, they required a period of learning to perform consistently. Once the necessary relationships between the tool and goal object were not physically situated in the task layout, as in true tool-use manipulation, the gibbons performed poorly. In a raking-in task where the necessary orientation for success had to be produced by the subject, no individual evidenced foresightful comprehension of the required action. There was some suggestion of learning the correct behaviour through associative processes. This finding was also supported by evidence from dipping experiments where the gibbons were provided with a transparent box containing a liquid reward and sticks that could be used as tools to access it. No individual developed dipping behaviour. The gibbons therefore, performed well on tasks when the salient relationships between tool and goal were directly perceivable. Once they became responsible for producing that relationship, performance was poor.

When the necessary orientation between the tool and goal was not provided by the experimenter, the gibbons evidenced low motivation to manipulate the objects. Given the gibbons’ requirement for direct visual feedback to comprehend the causal interactions between objects, this likely hindered their learning process. Failure therefore on the true tool-use tasks may not represent a particular cognitive limitation in these apes. A consistent finding was that the hoolock gibbons (Bunopithecus) were the most attentive and effective of the four gibbon genera. This is potentially due to the more variable natural environment experienced by these apes, driving selection for greater exploratory tendencies and flexibility of behaviour. The findings from this, and other work on primate cognition, suggest that contrary to propositions put forward by proponents of modular accounts of hominid brain evolution, the cognitive architecture of non-human primates contains neural mechanisms capable of processing technical information that may not be completely encapsulated. Suggestions that no non-human possesses specialised cognitive machinery for understanding objects as tools are also challenged.