Tracing crowds in social media writing

Social media stand accused of facilitating crowd-like behaviour, of sparking waves of angry messages in response to the latest media sensation instead of debate and critique. Such depictions draw on a distinction between publics and crowds, where publics are thought to be capable of reasoning, not least through the circulation of printed media, while crowds are understood in a derogatory sense as giving into affect and thus become easy targets for manipulation (Cody 2011, drawing on Tarde 1969[1901]). The new development brought about by the internet and social media is that crowds now seem to form not only through physical co-presence, but also in writing. For social research, this offers a chance to trace crowds in new ways and potentially come to question liberal assumptions about a clear distinction between publics and crowds. In this project, we attempt to trace and describe social media crowds in a massive data set consisting of Danish Facebook pages. First, we ask how crowds may be identified, and we experiment with different ways of locating imitation, drawing on Tarde’s idea that crowds are marked by a tendency towards homogeneity. Second, we ask what the implications are of tracing social media crowds: Are we crafting a detector of dynamics to be avoided in a liberal public sphere of rational debate, or are we perhaps rather discovering a contemporary version of the Cahiers de doléances – the descriptions of the grievances of ’the people’ collected right before the French Revolution – which may inform a more grounded understanding of the political terrain than the ’informed’ opinions and ’decent’ values of public debate (Latour 2017)?