Referential alarm calling behaviour in New World primates

There is relatively good evidence that non-human primates can communicate about objects and events in their environment in ways that allow recipients to draw inferences about the nature of the event experienced by the signaller. In some species, there is also evidence that the basic semantic units are not individual calls, but call sequences and the combinations generated by them. These two findings are relevant to theories pertaining to the origins of human language because of the resemblances of these phenomena with linguistic reference and syntactic organisation. Until recently, however, most research efforts on the primate origins of human language have involved Old World species with comparatively few systematic studies on New World monkeys, which has prevented insights into the deeper phylogenetic roots and evolutionary origins of language-relevant capacities. To address this, we review the older primate literature and very recent evidence for functionally referential communication and call combinations in New World primates. Within the existing literature there is ample evidence in both Callitrichids and Ce-bids for acoustically distinct call variants given to external disturbances that are accompanied by distinct behavioural responses. A general pattern is that one call type is typically produced in response to a wide range of general disturbances, often on the ground but also including inter-group encounters, while another call type is produced in response to a much narrower range of aerial threats. This pattern is already described for Old World monkeys and Prosimians, suggesting an early evolutionary origin. Second, recent work with black-fronted titi monkeys has produced evidence for different alarm call sequences consisting of acoustically distinct call types. These sequences appear to encode several aspects of the predation event simultaneously, notably predator type and location. Since meaningful call sequences have already been described in Old World primates, we suggest that basic combinatorial vocal communication has evolved in the primate lineage long before the advent of language. Moreover, it is possible that some of these communicative abilities have evolved even earlier, or independently, as there is comparable evidence in other taxonomic groups. We discuss these findings in an attempt to shed further light on the primate stock from which human language has arisen.