I’ve arrived at the University of Urbino for the inaugural AoIR Flashpoint Symposium, our new initiative that highlights important current issues in Internet research through one-day, concentrated events. This year’s symposium operates under the title “Under the Radar: Private Groups, Locked Platforms, and Ephemeral Contents.”
The first keynote at the AoIR Flashpoint Symposium is by Crystal Abidin, whose focus is on Internet celebrities. There are a number of different types of such celebrities, from well-recognised global celebrities to more niche micro-celebrities who are known mainly to a specific subculture. These people cannot be identified from their engagement metrics alone: they require a different set of research approaches.
Crystal builds on danah boyd’s idea of persistent, searchable, replicable, and scalable network publics here, but updates this to the concept of ‘refraction publics’: these are instead transient, discoverable by chance, decodable but not necessarily intelligible outside its original context, and exhibiting silosociability or rabbit-holing (where the intended visibility of the content is intensely communal and localised). In such refraction publics, human as well as machine activities are critical, contexts may be weaponised by interested groups, and there is a potential to alternate between public and private contexts.
One key aspect to this is visibility labour. There needs to be a useful amount of exposure and privacy in order to generate the right amount of visibility that is needed to reach and connect with the intended audiences. Such labour relates to an understanding of available communicative structures; to an appropriation of cultural practices; and to an engagement with platforms affordances to maintain agency.
Internet celebrity goes well beyond conventional celebrities; it also includes accidental celebrities who rise to visibility because of their embroilment in viral moments. Yet influencers represent a special category that has worked hard to attain a certain level of visibility and reach, and that reach ranges from all-visible mega-influencers to very nice nano-influencers.
Such influencers, their practices, and their self-understanding may be studied through digital ethnography studies, and the results will differ widely between different cultural contexts. It is also useful to track how (with the help of platform algorithms) certain influential contributors and their posts – for instance in reaction to major breaking news events – gradually emerge, and how a dominant narrative thus forms.
Crystal explores this in the context of Singapore – a highly technologised, highly surveilled society where alternative and underground media and activists have been forced to develop sophisticated communication strategies. One example of this are Internet celebrities who position themselves as consistently shameful by deliberately violating sociocultural mores, and use this notoriety to promote various commercial products. Another is the setting up of fake pages for official organisations (such as the public transport service SMRT) in order to troll them; this kind of ‘vigilante trolling’ is used to seek redress against perceived injustices.
\Some of this can be understood as subversive frivolity, intentionally remaining under the radar by maintaining the trappings of frivolity yet generating real impacts that subvert existing power structures. This is done in part also by gaming platform algorithms to gain greater visibility; by collaborating in groups to amplify each other’s posts (almost like a human bot network); by posting visual clickbait content (e.g. an image with a superimposed video play button); by posting looping GIFs that pretend to be live videos; by posting contentious content during the normally low-throughput ‘witching hour’ of 2-6 a.m. in order to attract new audiences and then delete them again after 6 a.m. in order to generate secondary searches; or by posting breadcrumbs and Easter eggs that hint at new stories.
Some groups also actively push out content in order to seed sentiment and generate new memes; they engage in fandom jacking in order to trigger fan groups into responding to and following them; they use Internet paralanguages and code-jacking to appropriate and adjust existing meanings; they participate in social steganography to rework existing content for local contexts; they engage in vague-booking or insta-vaguing by posting a stylish picture with an unrelated caption; or they use astroturfing to post what appear to be organic posts that nonetheless promote commercial products.