This thesis is about groups and their boundaries: how we bond with some people, but are separated from others who do not belong. It is also about social interaction - the building-blocks of this group identity. In particular, I investigate music’s role in our social landscape. Making music together is a powerful way of establishing and structuring these relationships; I argue that it can bring people together, but can also reinforce the divisions between them.

First, I present a new synthesis, drawing on relevant literature about our capacity for sociality, analyses of social interaction, and a history of the research on social groups. I outline a helpful framework by which to understand different forms of social engagement, depending on the nature of the interaction goal; I also clarify the concepts of interdependence and categorisation as distinct processes in group formation. Following this, I suggest that when our interaction is primarily affiliative, or relational, in goal (with little or no external goal focus), then it brings people together via relationships of interdependence; when we aim to communicate something more precisely (i.e. we have an external goal), then the need to maintain our common ground might instead form the basis for social division via categorisation.

Second, I report an initial empirical project which tests some of these predictions. My experiments show that music-making enhances affiliation, especially when there is no external goal focus. Adding a goal contributes to social division - affiliation on the basis of common team membership - but only when the interaction task was a success. When it was not successful, or caused embarrassment for those involved, participants instead seem to distance themselves from any associated group identity. This experimental work is supported by video analyses, in which I show different patterns of behaviour in interaction with and without an external goal.

This thesis is an important starting point in understanding the nature both of social groups and of music. It highlights the potential for music-making to structure our social world through either affiliation or division.