The Roots of the Ratchet: An Evaluation of the Capacity for Cumulative Culture in Non-human Primates

Abstract

Cumulative culture has been argued to be unique to humans, however, recent evidence from nonhuman animals suggests that relatively simple cases may be possible in other species. In this thesis, I explored whether non-human primates could strategically use information in a manner that may theoretically support cumulative cultural evolution. As the emergence of cumulative culture may depend on task demands, we hypothesised that favourable learning conditions may facilitate this effect. We presented monkeys with simple stimulus selection tasks, shown to produce evidence of cumulative cultural learning in human children. Success on these tasks depended on repeating rewarded and avoiding unrewarded stimuli, whose value had been revealed in a demonstration. Subjects were initially given task training to ensure that they had grasped the task contingencies and were required to meet a performance criterion before progressing. Subsequent test data was then used to simulate outcomes under linear transmission, using individual level task responses. This was done by aggregating the performance of individuals following multiple exposures to demonstrations of varying value. We tested whether, on average, exposure to higher quality information was associated with greater task success, a hallmark of cumulative culture, and whether subjects could improve upon the demonstrations they were presented with, representing increases in functionality. We also used the data to classify each subject’s potential for cumulative culture according to a continuum. Several individuals displayed the capacity for cumulative culture using this method, demonstrating that it is not precluded in non-human animals. However, this may depend on significant experience of relevant cues indicating the presence or absence of rewards (as provided for our subjects in the task training). Furthermore, increases in performance were limited, apparently due to relatively low precision use of the demonstrated information. Overall, this thesis suggests that, in non-human primates, cumulative culture might be limited to minor improvements accumulated over just a few episodes of transmission, and that it is likely to be rare due to its probable dependence on favourable learning conditions. This may help to explain the apparent rarity of this phenomenon in the natural behaviour of non-human species.