It has been suggested that a key function of music during its development and spread amongst human populations was its capacity to create and strengthen social bonds amongst interacting group members. However, the mechanisms by which this occurs have not been fully discussed. In this paper we review evidence supporting two thus far independently investigated mechanisms for this social bonding effect: self-other merging as a consequence of inter-personal synchrony, and the release of endorphins during exertive rhythmic activities including musical interaction. In general, self-other merging has been experimentally investigated using dyads, which provide limited insight into large-scale musical activities. Given that music can provide an external rhythmic framework that facilitates synchrony, explanations of social bonding during group musical activities should include reference to endorphins, which are released during synchronized exertive movements. Endorphins (and the endogenous opioid system (EOS) in general) are involved in social bonding across primate species, and are associated with a number of human social behaviors (e.g., laughter, synchronized sports), as well as musical activities (e.g., singing and dancing). Furthermore, passively listening to music engages the EOS, so here we suggest that both self-other merging and the EOS are important in the social bonding effects of music. In order to investigate possible interactions between these two mechanisms, future experiments should recreate ecologically valid examples of musical activities.

Group dancing is a ubiquitous human activity that involves exertive synchronized movement to music. It is hypothesized to play a role in social bonding, potentially via the release of endorphins, which are analgesic and reward-inducing, and have been implicated in primate social bonding. We used a 2 × 2 experimental design to examine effects of exertion and synchrony on bonding. Both demonstrated significant independent positive effects on pain threshold (a proxy for endorphin activation) and in-group bonding. This suggests that dance which involves both exertive and synchronized movement may be an effective group bonding activity.

It has been proposed that singing evolved to facilitate social cohesion. However, it remains unclear whether bonding arises out of properties intrinsic to singing or whether any social engagement can have a similar effect. Furthermore, previous research has used one-off singing sessions without exploring the emergence of social bonding over time. In this semi-naturalistic study, we followed newly formed singing and non-singing (crafts or creative writing) adult education classes over seven months. Participants rated their closeness to their group and their affect, and were given a proxy measure of endorphin release, before and after their class, at three timepoints (months 1, 3 and 7). We show that although singers and non-singers felt equally connected by timepoint 3, singers experienced much faster bonding: singers demonstrated a significantly greater increase in closeness at timepoint 1, but the more gradual increase shown by non-singers caught up over time. This represents the first evidence for an ‘ice-breaker effect’ of singing in promoting fast cohesion between unfamiliar individuals, which bypasses the need for personal knowledge of group members gained through prolonged interaction. We argue that singing may have evolved to quickly bond large human groups of relative strangers, potentially through encouraging willingness to coordinate by enhancing positive affect.