Search for a command to run...
Human language is probably the most complex communication system in the living world. It is investigated by various scientific disciplines, including linguistics, neuroscience, or cultural studies. However, despite this large and interdisciplinary effort, one key question has remained open and continues to perplex the scientific community; how could such an intricate system evolve? Comparative research on our extant evolutionary neighbours—the non-human primates—is often considered a good starting point to investigate the origins and evolution of human language. As humans communicate mainly with speech, primate vocal behaviour is the natural target of investigation, although this approach is not uniformly accepted. Behaviourist theories, in particular, characterise primate vocal behaviour as a predominantly hard-wired system, arguing that not much can be learned from it regarding language evolution. On the other hand, there is growing evidence for a considerable cognitive component in non-human primate communication, which often points to early signs of flexibility and indications of gradual evolutionary patterns more generally. In this thesis, I ventured to further our understanding of the flexibility in non-human primate communication systems through series of field experiments on wild South African vervet monkeys. First, I assessed the capacity of vervet monkeys to socially learn novel call-context associations. Using unfamiliar animal models in conjunction with alarm call playbacks, I showed that monkeys rapidly associated alarm calls with these models, evidenced by high vigilance towards them in the subsequent encounter. Furthermore, some juveniles also produced alarm calls similar to the playbacks they heard during the first encounter, showing how rapid social learning could influence call comprehension and usage in this species. In a second experiment, I tested the functional flexibility of vocalisations by providing wild vervet monkeys with opportunities to socially learn a novel usage of move-grunts to obtain food rewards. I worked with two groups that differed in the complexity of the learning stimuli provided during the experiments. For the first group of monkeys, I paired playback of movegrunt with a food dispenser providing a reward, such that the call predicted food as a simple conditioning stimulus. In the second group, I provided subjects with a demonstration video of a conspecific producing a move-grunt in order to activate the food dispenser and obtain a reward. While I did not find any evidence for relevant learning in the first group, a juvenile female from the second group started to produce her own move-grunts to obtain food rewards, suggesting that primates can learn to produce calls in completely novel circumstances if provided with the right social input. The focus of the third experiment was on flexibility in call perception. I examined whether vervet monkeys, when confronted with referentially ambiguous calls, use contextual information to respond to them. I addressed this by probing them with male 'leopard' alarm calls, which can naturally be given to terrestrial predators or during intergroup encounters. In the experiment, I played back leopard alarms either during natural between-group encounters or in a control situation. The subjects showed anti-predator responses and looked for additional information in control but not inter-group situations, suggesting that call meaning in primate communication is subject to simple forms of pragmatic inference. The results of these experiments indicate that non-human primate vocal communication rests on a primitive cognitive infrastructure that, within the human lineage, could have gradually evolved into the complex communication system seen in today's human languages. Finally, I conclude the thesis by proposing a classification scheme for non-human primate vocalisations, based on differences in underlying cognitive complexity, and briefly speculate about the future of primate research in light of emerging technologies that have the potential to revolutionise our understanding of the evolution of human and non-human communication.