| dc.description.abstracteng | It is difficult to overestimate the influence of language in day to day life, and language is often proposed as the clearest feature to distinguish us, Homo sapiens, from all other animals. The origin of language has long been the subject of speculation and myth, and more recently, together with the development of Darwinian evolutionary concepts, the evolution of language has become the focus of multidisciplinary scientific investigation. The emerging picture is that the language faculty is a complex multi-component system, with the possibility of each subcomponent having arisen from a different evolutionary pathway. Comparative studies of communication and cognition in non-human animals have the potential to identify whether these subcomponents, or biological precursors to them, are shared with other species, offering insights into the communicative abilities of our pre-linguistic ancestors and the evolutionary pre-cursors to language. To date there is no comprehensive list of what constitutes the language faculty; nevertheless vocal flexibility, semantic meaning and pragmatic inference are all fundamental characteristics of language and comparative studies can thus be used to shed light on how these particular features arose.
In this thesis I investigate vocal communication in the African green monkey (Chlorocebus), with especial focus on alarm calls. The African green monkey is a widespread genus consisting of six species, one of which is the vervet monkey (C. pygerythrus). Numerous studies have been carried out on vocal communication in the vervet, and their alarm calling system is the classic example of functional reference (that is, calls that function in a referential manner) in a non-human species. Nevertheless there are still many open questions regarding the mechanisms underlying call production and call perception in this species, and very little is known about vocal variability across the genus. In three inter-related studies I therefore explore questions of vocal flexibility with respect to the degree of control over vocal production, questions of semanticity concerning the relation between vocal signals and external stimuli, and questions of pragmatics pertaining to whether signal receivers incorporate contextual cues when responding to conspecific calls. The overall aim of these studies is to increase understanding of the proximate mechanisms and selective pressures that shape the production and perception of functionally referential alarm calls in the African green monkey, with the broader goal of offering insights into the emergence of vocal flexibility and meaning in language.
To investigate vocal flexibility in the adult male alarm bark, I carried out a study to assess geographic variability in vocal structure within and between African green monkey species, and the function of any structural differences. To do this I recorded alarm barks from male green monkeys (C. sabaeus) in Senegal, and from male vervets (C. p. pygerythrus) in South Africa, and compared these to alarm barks made available to me from East African male vervets (C. p. hilgerti). I then carried out playback experiments in which the barks of an own-group male, an unknown conspecific male and unknown heterospecific (but congeneric) male were broadcast to adult male South African vervets to test whether they would distinguish between them. All barks demonstrated a similar overall call structure; nevertheless, inter-specific differences in fine-scale spectral and temporal characteristics allowed green monkey barks to be distinguished from vervet barks with a high degree of accuracy. East African and South African vervet barks also exhibited spectral differences in call structure, although intra-specific differences were smaller than inter-specific differences. In response to playback experiments, males typically reacted most strongly to unknown conspecific barks, and showed behaviours typical of leopard-avoidance and male-male aggression. The findings of this study thus imply that the acoustic structure of male barks is innate, and that variation is likely related to phylogenetic differences. Furthermore, barks may have a dual function as both an alarm and display call, offering a potential explanation for the sexual dimorphism observed in Chlorocebus alarm calls.
The seminal finding that vervet monkeys produce predator-specific alarm calls that elicit predator-appropriate response behaviours led initially to the conclusion that these calls are semantic signals akin to human words. The inferences that can be drawn regarding a call’s semantic value based only on the responses it evokes in others are, however, limited. In a second study, I revisit the vervet alarm calling system. Using recordings made available from the original studies together with additional calls I recorded myself, I carried out quantitative acoustic analyses to assess the degree to which calls are structurally discrete and context specific, both classic criteria for functional reference. These acoustic analyses reveal that at the level of the call element, vervet alarm calls are distinguishable with regards to the predator type eliciting calling, but that differences in spectral structure are rather graded; especially in the case of male calls. When compared with calls produced during social and aggressive interactions, female and male alarm call elements are less distinct. Importantly, temporal characteristics of call series distinguished between spectrally similar calls produced in different contexts. Thus at the level of the call element, vervet alarm calls constitute a graded system, especially in males, and context-specificity is reduced when calls from non-predator contexts are also considered. The adaptive value of distinguishable alarm calls in this genus, may have led to the production of temporally distinctive call series.
Green monkey females produce acoustically similar alarm chirps to snake and leopard models. In study three I 1) quantified acoustic differences between snake and leopard chirps, 2) used predator models to identify typical response behaviours to snakes and leopards, and 3) conducted playback experiments to assess whether subjects would respond to chirps with predator-typical response behaviours, and whether contextual cues in the form of pre-exposure to a leopard or snake model would influence these responses. Chirp calls exhibited graded differences between leopard and snake contexts; nevertheless, subjects responded to leopard chirps with leopard-appropriate behaviour irrespective of contextual cues. Interestingly, priming condition did not have an effect on the subjects' immediate response, but, together with call type, did influence subsequent behaviour.
In conclusion, the observed variation in vocal structure across species of the African green monkey is likely related to phylogenetic differences, with more flexibility in temporal than spectral characteristics. In keeping with findings from all other primate taxa to date, this is indicative of a largely innate vocal production process with limited control over call structure. At the level of the call element, vervet alarm calls show considerable intergradation, both among different alarm call types and between alarm and non-alarm calls, a finding in line with neurological studies that suggest that structural features of primate calls are strongly influenced by the caller's motivational state. Considering call perception, acoustic and contextual cues influence green monkey responses to a graded alarm call. Taken together, the findings of these studies suggest that functionally referential signalling systems provide little insight, on the side of the signaller, into the evolution of flexible and arbitrary symbolic communication. More relevant questions are how cognitive and emotional factors interact to influence call production, and how contextual cues are integrated into processes of call perception in non-human animals. Such comparative research is highly relevant to studies of emotional prosody and pragmatic inference in humans, with implications for a better understanding of how language evolved from the innate and emotionally grounded communication system of our pre-linguistic ancestors. | de |