The Politics of Trending

“Twitter determines its trending topics through the site’s Trends algorithm, which Twitter has not released to the public.”

I’m not afraid, Free Trade!

–Don Mee Choi

Every time CNN points to “Trending” in order to discuss breaking news, we should laugh. That is, laugh at: CNN, journalists, experts, the simulacra. Smirk at the notion that privatized, opaque institutions of selective coverage are working with other privatized, opaque institutions of selective timelines to define what’s public, what’s universal, what’s important.

Intellectual, digital and digitized labor is important, and too often dismissed. This interrogation of visibility, coverage and trending isn’t a critique of online activism and discourse. Rather, I want to examine the political pontification around “trending” and “tagging” — the politics of Twitter algorithms, the bird’s eye commentary of “surprise” and “dismay” at certain trends to ultimately locate the splinters of alternative discourses.

The Crisis of Visibility

Photo CC-BY Lee Rui Kang, cropped and filtered.

#Solidarityisforwhitewomen was a hashtag created by Mikki Kendall in August of 2013. The tag presented a structural critique of global north feminism and the oppressive dynamics underlying white feminist understandings of “empowerment” and “progress.” The tag was also a direct critique against white feminists who defended “male feminist” Hugo Schwyzer. The hashtag trended worldwide, prompting a fury of articles by supporters, surprised journalists and critical commentators.

Criticism of tags like #solidarityisforwhitewomen situated the trending of such conversations as an exceptional and at the same time misdirected, mis-use of digital energy by racialized and gendered users. However, hashtags like #happybirthdaytaylorswift and #happybirthdaydemilevato have yet to receive “critical” scholarship, essays, exposes and op-eds. This is because their themes fit into the dynamic and expectations of what is expectedly visible, what a trend might constitute, and what cyborgs are supposed to be interested in.

Michelle Goldberg’s “Feminism’s Toxic Twitter Wars” focused on criticizing the digital labor of black feminists online, particularly Mikki Kendall who created #solidarityisforwhitewomen. The crux of her argument focused on how surprised mainstream feminists were by critiques of their objectives, agendas and methodologies. Goldberg cites Kendall as the originator of #solidarityisforwhitewomen and as one of the main critical voices of the #femfuture tag, which came under scrutiny for its blatant WoC erasure, questionable twitterethics, flattening of difference, and concerns surrounding accessibility. Goldberg defended that “#Femfuture was earnest and studiously politically correct” and that organizers were “floored” to learn that feminists online were critiquing their efforts. How dare feminists disagree with #femfuture, and then go on to trend a movement of their own? Before addressing the claims of the critiques, Goldberg states that such conversations create a ‘toxic’ environment for feminism online.

Goldberg’s analysis conveniently ignores how #solidarityisforwhitewomen and criticism of #femfuture were formed to critique white supremacy in feminism (a historical, ongoing, structural analysis). Visibility seems to be the central concern for Goldberg — not the critique of white supremacy within feminism — but the visibility of such critique. Goldberg’s attachment to the nature of such tags is that they are exceptional, singular;  in fact, her concern for visibility might be better interpreted as a PR concern. In her view, critiques of white supremacy within feminism are exceptional concerns, so they should receive limited visibility. However, they have received exceptional attention — and this, and only this, must be rectified.

The Management/Displacement of Structural Critiques

Photo CC-BY Sean Freese, filtered.

While visibility might be a central focus for white feminists like Goldberg, visibility is not always the central objective for trans, women of color, and black feminisms. At the “We Cannot Live Without Our Lives: A Conversation on Anti-blackness, Trans Resistance and Prison Abolition” forum at University of California, San Diego, activist and artist Reina Gossett problematized the function of visibility for black trans women. Gossett articulated that particularly for black communities, “Visibility is a pillar of criminalization, not a tenant of liberation.” In conjunction, Grace Hong has argued in The Ruptures of American Capital that “[F]or women of color feminist practice, visibility is a rupture, an impossible articulation.” Hong writes that while some have articulated invisibility as unnatural, “[S]o too, is visibility is unnatural; it is also a kind of violence…visibility is not inclusion, but surveillance.” The operations of structure cannot be wholly represented. The representations of structure are also its contortions: there are many reasons to resist representation and inclusion.

Visibility — while perhaps essential in the grab for legitimized forms of violence and power (state power, representational power, corporate power) — remains one condition of the expressions of structure. What’s visible is crucial because it’s the surface representation of structure. But as Gossett and Hong have pointed out, to be exposed and figured in the surface has its own severe limits. Critics like Goldberg fixate on ‘what has become visible’ to protect representations linked to the privileges of the status quo, rather than tend to the ongoing damage of structural violence.

We see this refusal to engage the structural across a breadth of trending topics. #Gamergate is/was a movement entirely unlike #solidarityisforwhitewomen. In fact, they are of polar opposite camps, brought forth by entirely differing subject positions. Formed-ish in conjunction with multiple conversations that were happening on Reddit and other channels for online communique, and trended more or less by actor Adam Baldwin’s tweets, the overarching claims of Gamergate were “criticisms” (terrorism) of game developer Zoe Quinn and feminist critic Anita Sarkeesian, eventually expanding to target many other women across gaming and tech. The heart of the issue from the beginning was: who gets to criticize, and who does not.

The response to #Gamergate invokes structural critiques of the ongoing, well-documented culture of misogyny, rape and patriarchy in tech and gaming. Yet rather than grappling with the structural violence #Gamergate reflects and the active resistance that it illuminates, many commentators across platforms jumped for damage control. Gaming executives expressed dismay that harassment was “tarnishing our reputation as gamers,” while academics bemoaned: “We have been working for years to make games a legitimate tool for education and for study, and we were making progress… And then came GamerGate… now, when I go to talk about games to industry groups or fellow academics, GamerGate always comes up as an example of how terrible and immature people who play games are… It will take years and years to repair the damage…”

Here, reactions of ‘surprise and dismay’ eschew structural engagement; those that disapprove or are embarrassed by the trend seem more to have an issue with the visibility of the problem, concerned more with “damage control” “image” and “PR” rather than care for the damage, care for the issue and structural adjustments.

The Algorithms of Visibility

Photo CC-BY Jared Tarbell, filtered.

Panic driven by visibility is predictable. Zeynep Tufekci has argued that too often digital humanities or scholarship around the digital “[R]arely goes beyond exploring big data as a hot, new topic and an exciting new tool, and rarely consider issues of power.” The “analyst,” in focusing in on the function of technology, of trending, completely forgets to discuss the structures and dynamics of power materialized every step of the way by this tech, the data and its users. And so it holds that while trends are provoking of such dismay and surprise, are so formidable, so worthy of journalistic inquiry and coverage, there is little to no critical analysis of “trending” itself. What is the material, what are the logics and the substance of trending when:

1. We don’t know why something trends. The algorithm is a locked secret, a “black box” (to the point where MIT professors have built algorithms attempting to predict trending tags). The Fineprint: Trending is visibility granted by a closed, private corporation and their proprietary algorithms. As Tufekci says, “Algorithms have consequences.”
2. The visible trending box is supposed to vary according to personal preference. There are algorithms for localized trends, “neutral” US trends, global trends, and other a la carte options. The Fineprint: The Algorithms can and should be adjusted according to personal preference — we want our reach to be individualized.
3. The little bit of information the private developers have released is that a ‘trend’ is based on a very specific definition of ‘now’ and ‘new,’ a definition that we as users do not have access to. The Fineprint: something cannot trend for too long, this isn’t their definition of a ‘now’ and ‘new’. This is why Ferguson failed to trend after a few days even though it was one of the most widely used-hashtags — trending for a few days excluded it from the possibility of trending.
4. Concerns about why certain HTs don’t trend (ie #occupywallstreet #wikileaks or the various other #occupy’s) will lead Twitter developers to tell you that perhaps something is not as popular as you think it is. The Fineprint: trending is what they believe is popular, a paradoxical assertion: private formulas declaring what is most public and ‘new’.

Through this lens, trending is merely visibility granted by the algorithms of a closed, private corporation:

Trends: what is happening, new, now according to Twitter the corporation’s ever-changing algorithms

Trends: a moment in which the “public” space of the internet becomes concentrated

Trends: a concentration of trolls

Trends: concentrated energy

Trends: visibility granted by a closed, private corporation

Trends: the commons managed by a closed, private corporation

Trends: manifestation of algorithms

Trends: expected

Trends: unexpected  

To further illustrate how much we don’t understand why something trends, I used Topsy to provide me with analytics on the usage of #FreePalestine, and compared it to the various trendings tags on March 1st, 2015. This experiment was prompted by users throughout last summer and this year, observing how #KillAllMuslims trended recently but #FreePalestine has been unable to trend.

As of January 30th to March 1st, #FreePalestine has been utilized over 84,333 times.

Yet March 1st’s United States Trends were:

So how might #WeWantTheCup fare next to #FreePalestine?

That’s 767, total. Not 767,000 shortened.

Or #ExplainAMovieByItsTitle:

or #KillAllMuslims?

How might all of this fare next to the usage of #Ferguson?

#Ferguson has been used over 2,000 times on March 1st alone… clearly surpassing the usage of the current US trends of #WeWantTheCup and contending with the usage of #ExplainAMovieByItsTitle. Basically, #Ferguson should be trending everyday.

Of course this Topsy graph is limited to Jan 30th-March 1st, so all of these hashtags could’ve been utilized differently the months and years before. And of course, Twitter developers have already graciously explained that trending is their particular definition of “new” and “now” and so previous tags, or tags used continuously (such as Justin Beiber tags, which have been implicitly banned from trending) will not trend.

I bring this up to point to how the “journalistic scholarship” around visibility and trending is completely and utterly misinformed, misframed and just plain silly. The exceptional attention given to hashtag discourse by critics, news platforms and journalists — to what they perceive to be evidence of visibility — takes the focus away from the spaces created by gendered and racialized users, and rewrites it as a singular confrontation racialized/gendered users are having with white audiences within a white space. This rewriting positions trending tags to be isolated explosions. It does not labor through the possibility of communal, ongoing engagement and sustainment, for better or for worse. Though this is clearly their fixation, this fixation should not prevent us from thinking through and recentering the persistent and ongoing labors involving disobedience, disturbance and cyborg mutations: alternative discourses.

Hacking a Commons

Photo via Electronic Disturbance Theater, used here with permission.

Rather than fringes, minor and eclipsing, hashtags are digitalized fragments of political ancestry; ongoing, replenished, connected to and beyond their current framework — ancestral, rather than marginal (this is a poetic framework I’m borrowing from Lucas de Lima). The trending of #solidarityisforwhitewomen #ferguson #blacklivesmatter #mediablackout, and the usage of tags like #FreePalestine demonstrate that radical conversations are not exploding or momentary: they are ongoing and variegated. They are leaps of politicalized conversations in what was supposed to be the apolitical, private sphere of Twitter.

Indeed, Twitter is not designed as public even as it fundamentally derives from public input and data, and parades as common grounds. Luis Martín-Cabrera offers that, “Karl Marx saw the appropriation of the commons as one of the elements of  ‘primitive accumulation,’ a ‘ground zero’ of ‘surplus value.’” That is, the shifting of the public, and the taking of the commons is the basis of privatization — a process that exists to exploit the majority for a minor few. However, you don’t have to be a Marxist to follow the argument that: privatization is antagonistic to notions of the commons, and explicitly closing the definition of “the public” to the public is manipulative and deceitful. Trending highlights only what they want us to see, what they’re allowing us to see. Fixating on trending/visibility is the secondary layer of their gaze.

But there are some tools being utilized to circumvent Twitter’s original design, to disobey aggressively, to claim Twitter as a true “commons” — a space marked for public debate and protest. This commons is bigger than their grasp, linked in uncontrollable ways, and subject to reclamation, to subversion… to hacking. #Gamergate as a tag seamlessly fit into and inside sets of opaque algorithms, into the discourse of public trending — rooted in consumer activism, misogyny and violent calls to preserve patriarchy. However, criticism of #gamergate, similar to criticisms of #femfuture, shifted and subverted the hashtag’s primary narrative. The trending of #solidarityisforwhitewomen was a hack of opaque, proprietary algorithms. Rather than what is being granted visibility, I am interested in these, and related efforts of affirmative, intimate sabotage:

• the communities that congregate around tags, regardless of their trends
• the users that notice when important topic (such as #Ferguson) are not trending, and use alternative tags (such as #mediablackout)
• users who pre-emptively create “when in jail” accounts (planning the heist!!!)
• anti-doxing collectives
• users on Twitter & Tumblr, that retweet/retumble posts but tag their critique (for example, this is not native looks through various “native” “indian” tags and retags the “original” with links on appropriation, genocide, and add: appropriation.)
• and so much more

This is the ongoing praxis that actively questions the stakes of power in our privatized digital publics. Intimate sabotage: In becoming close to the opaque, being fed by it yet reproducing otherwise. Refusing replication: becoming its expert and traitor. Learning everything about it—disenchanted of its awe, seamlessly wandering inside of it, finding novel ways of attack.

Intimate sabotage of Twitter is a provocation: that any sense of the public in this privatized milieu will have to be reimagined. In discussing the potentiality of the commons, Martín-Cabrera extends, “[T]he potentiality of the commons can only be actualized when we actively disobey and when we actively ‘connect and fight.’” What is the baseline for a commons as fighting ground? The idea of unmanageable as the commons: one that consistently erupts with conversations about afro-futurity, race, gender, feminism, colonialism and the police state. The deployment of intimate sabotage as communal building: the algorithms hacked, broken, reworked for purposes beyond the immediate reach of private formulas, private means.

Tufekci argues that we continue to live with ineffective models/tools against our oppression; she suggests that “We need to update our nightmares.” The nightmare before may have been — some of us are invisible — before such concerns can be remedied — the update indicates that invisibility can no longer be exchanged evenly for represented visibility. Other updates may be: our visibility is our surveillance (Hong), our visibility must be contested (Gossett), our visibility is not very particular — and we must find the tools to sabotage it.