This research uses computational methods to analyse historical interactions about religion on Usenet, an early online social networking service that prefigured modern social media. It uses an archive of messages from c.1981-1991 and later archives from c.2003-2015. I define and analyse how cohesion and incivility manifest in the interactions, the extents to which they occur, and factors which are associated with these social processes.
The research is designed at a multidisciplinary intersection of the study of religions, Internet history, and computational methods. It is foremost religious studies research with a particular focus on methodology and digital networks. The study of religions provides the concepts of community and authority, to which I relate my findings, along with insights about how digital interactions are part of broader personal and social lives through which religion is expressed and experienced. Internet history provides technical detail about early computer networks and the social and scientific contexts in which they emerged. This ‘digital substrate’, as I term it, shapes how participants engage on Usenet. Internet history also provides methodological ideas and critical evaluation about how archives of online interactions can be used for research. Computational methods — here, the language R — enable me to form ethically sensitive datasets from archives, identify occurrences of social processes, and illuminate associations between facets of the dataset. For example, this includes identifying differing patterns of participation across online social networks (discussion groups) within the sites researched.
I find that computational methods are useful for large scale analyses so long as researchers understand how the computer code in those methods functions and they mitigate for imprecision. The methods are suited to identifying general trends, patterns, and associations, which can then be triangulated through manual analyses. I specify ways in which cohesion and incivility do indeed manifest, some factors associated with them, and how these inform academic understandings of community and authority when people connect online to discuss or practise religion.
I conclude that the sort of methodology used and the ethical approach embed within it are suited to analysing a broader range of social processes in the same and other similar historical online contexts. I also argue that while Usenet has been surpassed by other social media, its archives remain relevant since (1) they capture interactions spanning four decades of history, and (2) the affordances of historical online social networking, and the ways in which people behaved, are relatable to current online text-based social networking environments.
