Researchers often know very little about social media users other than their public actions on a platform (for example, posting, sharing, liking) and when these actions occurred. In this sense, big data has been described as “thin” in that it lacks information on social context and the motivations and beliefs of actors. This makes it challenging to characterise the identity of social media users, which is important for understanding, for example, influence and communicative patterns of actors in digital spaces. In this research we present a new approach for representing identity in online networks, drawing on two distinct bodies of literature.

In Social Network Analysis (SNA), identity is often conceptualised as pre-defined actor (node) attributes, such such as age, gender and education. The relationship between network structure and identity is therefore often treated as static and fixed, but this ignores the fact that identity is continuously enacted, performed, and reforged through social practices and social interactions: identity is not separate to network structure but co-constituted by it. However a significant strand of SNA research does focus on how associational data can be used to understand identity: Ron Breiger, in his foundational work on the duality of persons and groups, remarks “...the particular patterning of an individual’s affiliations defines his points of reference and (at least partially) determines his individuality” (Breiger 1974, p. 181).

The second body of literature we engage with begins with the sociologist Gabriel Tarde who more than a century ago devised a radically new way of conceptualising and studying the social world. Drawing on Leibniz’s notion of the ‘monad’ (Tarde 1983), Tarde set out his monadological theory, but precisely what Tarde meant by a monad remains debated in the literature, despite its further development within the contemporary field of Actor-Network Theory (ANT) (e.g. Latour et al. 2012).

In the present research we draw on both SNA and ANT to propose four axioms for representing identity in networks. We demonstrate that a benchmark network science approach only satisfies one of these axioms, while our new monadological network approach satisfies all four axioms and provides qualitatively different insights in terms of the importance and influence of actors. We demonstrate our approach using a large-scale Twitter dataset related to the 2023 Australian Referendum on the Indigenous Voice. While our focus is social media networks, our approach can be used in any setting where there is dynamic affiliation data, for example, people and the events they attend over time.

Professor Robert Ackland holds a PhD in economics and is a professor in the School of Sociology at the Australian National University (ANU), specialising in social network analysis, computational social science and the social science of the Internet. Prior to commencing his academic career, Robert worked as a researcher in the Department of Immigration and the World Bank. Robert has been studying online social and organisational behaviour since 2002, and he leads the Virtual Observatory for the Study of Online Networks (VOSON) Lab (http://vosonlab.net) which he established in 2005 under an ARC Special Research Initiative (e-Research) grant. His recent areas of interest include network approaches to studying online political deliberation, echo chambers, misinformation and the role and impact of social bots. His book Web Social Science: Concepts, Data and Tools for Social Scientists in the Digital Age was published by Sage in 2013

References

Breiger, Ronald L. (1974), “The Duality of Persons and Groups,” Social Forces, 53(2):181-190.

Tarde, G. (1893, 2012). Monadology and Sociology. re.press: Melbourne.

Latour, B., Jensen, P., Venturini, T., Grauwin, S., and Boullier, D. (2012), “‘The whole is always smaller than its parts’: A digital test of Gabriel Tardes’ monads,” British Journal of Sociology, 63(4):590–615.