Machine learning reveals adaptive maternal responses to infant distress calls in wild chimpanzees

Abstract

Distress calls are an acoustically variable group of vocalizations ubiquitous in mammals and other animals. Their presumed function is to recruit help, but it is uncertain whether this is mediated by listeners extracting the nature of the disturbance from calls. To address this, we used machine learning to analyse distress calls produced by wild infant chimpanzees. It enabled us to classify calls and examine them in relation to the external event triggering them and the distance to the intended receiver, the mother. In further steps, we tested whether the acoustic variants produced by infants predicted maternal responses. Our results demonstrated that, although infant chimpanzee distress calls were highly graded, they conveyed information about discrete events, which in turn guided maternal parenting decisions. We discuss these findings in light of one the most vexing problems in communication theory, the evolution of vocal flexibility in the human lineage.