#notracist: Multi-Hashtags and Ambient Race Talk on Twitter

Twitter is becoming known for the abundance of racialised messages posted on its platform. Eruptions of racist abuse occur within a contested array of Twitter discourses, e.g. racial banter, ambivalent humour, hate comments, trolling and (anti-)racist sentiments. While there been some focus on high-profile racialised controversies and events on social media platforms, this attention can effectively ignore more mundane, yet racially charged online talk. Surprisingly, a study of everyday ‘race-talk’ on Twitter has not been undertaken, and little is known about its ambient stream of racialised expression.

This paper explores how racialised messages unfold in Twitter by focusing on the hashtag #notracist. Approximately 25,000 tweets which included this hashtag were collected over an eight month period. Hashtags enable a Twitter message (tweet) to become more findable and the potential to join a larger conversation or ad-hoc public. A methodology was developed using data visualisations to examine the corpus of Twitter data, which was not based on any specific event. Twitter users can include #notracist (and other related) hashtags in messages to label seemingly ‘racist’ statements (or links and images/videos) as ‘not racist’. In particular, the practice of using more than one hashtag (multi-hashtagging) – especially in the topical forms of either ‘humour’ or apparent ‘factual observations’– is found to be a key strategy of denying racist expression and propagating the ambiguities of Twitter race talk. Moreover, it is discovered that hashtags can be used as disclaimers, rather than for eliciting conversation. In contrast to focussing on, or interpreting the meaning of tweets, this paper explores how hashtags– as digital assemblages – are the message, and generate different practices of racialised tweeting. Overall, this paper offers a methodologically informed study to critically understand the contested digital ecology of Twitter.