DOES CHAOS MAKE MAMMALIAN CALLS SOUND MORE ALARMING TO HUMAN LISTENERS?

Mammalian vocalisations are extremely diverse, and evolutionary approaches to understanding this diversity assume that much of their acoustic form maps onto their function, with specific features serving universal roles. Here, we hypothesize that nonlinear phenomena (i.e., deterministic chaos, sidebands, subharmonics, and frequency jumps), which make the voice perceptually rough or harsh, contribute to the alarming quality of mammalian vocalisations. To test this, we investigated whether adding such acoustic irregularities to synthetic replicas of calls produced by a range of terrestrial mammals increased their alarming quality. We found that, independently of the species, the presence of nonlinear phenomena, especially chaos, made calls more alarming for human listeners. While these results hint at a universal function of NLP in the mammalian vocal repertoire, future studies should now investigate whether the alarming effect of nonlinear phenomena holds for non-human species.