My research concerns the mechanisms, functions, and evolutionary basis of nonlinguistic vocal communication, such as human laughter and nonhuman primate calls. The crux of the approach is that vocal signaling in mammals has its origins in vocalizers using sound to induce affective responses in listeners, meaning that the vocalizations "work" by influencing attentional, arousal, emotional, and motivational states in these perceivers. This view was developed in collaboration with field primatologist Drew Rendall (University of Lethbridge, Alberta, CA), and has been dubbed the "affect-induction model." We contrast our approach with the predominant view that nonlinguistic vocal (and other) signal) are inherently representational--meaning that when signaler communicate to perceivers, they are conveying encoded information about either their own affective state or some exteral object or event.

We argue that human and primate vocalizers may be relying on two much simpler strategies. First is that socially impotent individuals (like infants, young offspring, or vocalizers that are socially subordinate to the listener) tend to rely on the "direct" impact of the acoustics of the signal itself. Young vocalizers, for example, use repeated high-imact sounds like crying, shrieking, or other kinds of loud, "extravagant" vocalizations as a crude way of drawing the attention of listeners, inducing response motivation in if these perceivers are reluctant to act, and even in effect training caregivers in how to meet their needs in order to "turn off" the potentially noxious sounds being produced.

Socially potent individuals, on the other hand, can use a very different, but nonetheless almost equally simple tactic. This "indirect" strategy is to produce sounds that are paired with other affect-inducing actions or circumstances. If those sounds are rich in cues to the signaler's identity, the listener will requisitely acquire conditioned affective responses to that particular vocalizer's signals. For example, if a dominant monkey produces either a threat call that is followed by attack or a contact call that is followed by a grooming episode, the targeted subordinate listener will quickly learn either fearful or positive anticipatory associations to these sounds. In humans, we propose that laughter produced with friends in happy circumstances is being paired with positive affective affect states in those listeners. When laughter occurs consistently and reciprocally in such settings, a listener acquires learned positive reponses to the vocalizer's distinctive laugh sounds. Occurring reciprocally, laugh production becomes a mechanism of fostering cooperative behavior by creating mutually positive emotional stances, and can be used as affective leverage that helps ensure friendly treatment when conflicts arise.

Empirical work testing these ideas has emphasized production and perception of species-typical calls in nonhuman primates, domestic cat meowing to human listeners, attentional and arousal responses of human infants to the exaggerated acoustics of infant-directed speech, and the acoustics and perception of human laughter. The laughter work has been conducted in collaboration with Jo-Anne Bachorowski (Vanderbilt University, Nashville, TN), where the affect-induction model provides a framework for understanding audience-dependent variation in laugh rate and acoustics, as well as differentiated listener emotional responses to laugh sounds.