This is an old revision of the document!
This project exploring the the #notracist dataset develops a practice-based view of Twitter. Our interest is not in using linguistic identifiers (such as individual words or collections of aggregated 'topics') to identify and classify types of racialized talk, but to probe how Twitter users do the work of racialized talk. This interest turns us towards a more methodological investigation of how researchers might recognise and depict different practices, techniques and modes of tweeting in terms of what they are, how they function, and how those practices contribute to the ‘achievement’ of a tweet. That is, how the #notracist (and additional) hashtags express a tweeter’s apparent intentions, and how racialized hashtags do the work of 'tweeting a joke' or an 'objective statement'.
Tweeting is not a restricted or flawed form of communication, but rather the result of the efforts of users who are well versed in Twitter as a medium and in the usage of Twitter as part of a wider ‘technocultural assemblage’.1) This view is particularly apposite to Twitter, given the artful, user-generated nature of much of Twitter's developing syntax. The most notable example of this is Twitter's hashtag function. Early on in Twitter's history, users began to coalesce on the idea of prefixing key topical terms with hashtags (# symbols) as an agreed-upon ‘folksomny’ method of categorising topics and making those topics searchable via Twitter's (then) rudimentary search function.2).
Our analytic orientation takes cues from a materialist perspective, which conceives Twitter as an assemblage. That is,
‘digital networks, communication platforms, software processes … are constitutive of online racialized subjectivity and activity. Pursuing a materialist approach … leads to re-conceiving race as an ‘assemblage’: encountering race as an emergent force in digital media vis-a-vis its networked connections, informatic flows and affects’3)
Much of the study of online racism tends to focus on highly visible & extreme forms. While this work is very important, nonetheless, it may miss out on grasping the materiality of social media and how this produces less visible, everyday online racism. Researching everyday, rather than extreme racism in social media, is methodologically challenging; not only because of the contested definitions of what constitutes racist-talk, but also, how to identify this on Twitter.
The method developed in our project focussed on the specific use of the hashtag #notracist, which we discovered captured certain kinds of everyday race-talk. A visual analytic approach was pursued based on utilizing Chorus. The figure below is a time-line representation of the #notracist dataset over the 8 month period of data collection. Chorus displays various metrics such as tweet frequency (light grey), links, and topical novelty. Essentially, we captured a steady, relatively low-volume of tweeting activity across a wide array of messages, which bubble away on Twitter, often without ever trending.
What’s significant about this time-line data visualization, is that the #notracist hashtag tends not to be about any specific event, issue or conversational topic. Another way thinking about this, is rather than highlighting popular/most active tweeting – which indicates some kind of controversy or 'extreme' event – instead the #notracist hashtag enables an exploration of the ‘long-tail’ racism of Twitter – the everyday racialized background (ambient) chatter.
The long-tail is produced by network effects – a power law distribution – which while commented upon in social media research, it's rarely studied in-depth. Paradoxically, it can be said that many social media users, including those on Twitter, spend most of their time residing in the banality of the long-tail.
It’s worth pointing out that analytically, this focus requires a different approach, than the time-dependent or event-based studies which dominate existing Twitter literature and valorize the highly visible and spectacular. Arguably, a ‘Twitter Studies’ is emerging primarily based on event analysis, which is caught within a ‘real-time presentism’ that lacks a wider frame of understanding.4)
In contrast, our study aims to explore how everyday racialized talk remains 'visibly hidden' and expresses modes of 'racial paranoia' - see Jackson (2008).5) After the era of political-correctness, racist talk is increasingly publicly unacceptable, and has been legislated against, as a form of hate-speech. Although, this doesn’t mean racist expression has disappeared. In fact, in some situations it’s become repressed, operating more underground in relation to its pubic unacceptability. Yet the virality of Twitter can produce astonishing explosions of ‘crowd-sourced’ racisms and abuse. When this type of racism does appear on social media it’s normatively understood as socially aberrant, and publicly perceived as ‘extreme’ in nature.
These types of explosive racist Twitter events in fact obscure how social relations continue to be governed by everyday forms of racism. As David Goldberg & Philomena Essed have maintained, expressions of racism and its simultaneous denial, is not an aberration, but quite ordinary and formative of western liberal societies. It’s been highlighted that strategies of the ‘denial of prejudice’ pervades everyday race talk.6) And practices of denial take the form of a disclaimer, such as “I’m not racist, but …”. In relation to Twitter, the inclusion of the hashtag #notracist echoes this strategy of denial. Arguably, #notracist materializes a racial paranoia that entangles 'privatized' expressions of racism, with 'public' modes of denial. Twitter is seemingly a unique for collapsing the private and public, which may explain why racialized talk is prolific on this platform.
<Back to project homepage>