What, Exactly, Was Twitter?

Twitter logo, c. 2012. By Martin Grasser, Wikimedia commons.

One of the perils of trying to theorise social media is that we often find ourselves talking about the status quo ante, the period immediately before. Twitter has always had a much smaller active user base than some of its social media rivals, such as Facebook. Nevertheless, it exercised an extraordinary influence over how politics and culture (and the politics of culture) were talked about during the 2010s.¹ Twitter became the chosen social media hangout for the intelligentsia, that fraction of society tasked with shaping worldviews and putting forward authoritative interpretations of the world, and it was the magnetic pull of that group of users that drew others to the platform.² The presence of “big names” and staff writers for The New Yorker attracted a wider penumbra of users who wanted to see what the intelligentsia had to say and, through the reply and the retweet, to be occasionally seen by the intelligentsia themselves. The novelties of interactivity and the default visibility of the public profile meant that readers could feel like they, too, were an intrinsic part of “the discourse.”

Part of what made Twitter so addictive for writers and humanities academics was that its character limits artificially concentrated speech and thus seemed to intensify its power. By reducing the average tweet size to the length of a Dorothy Parker-style bon mot, it gave those with the ability to condense their thoughts into the “take,” the witty rejoinder, or the pithy assertion a new kind of authority. More generally, it was also a place where one could make one’s writing available instantly and almost effortlessly to a huge potential audience—that fuzzy but nonetheless swarming penumbra of interested readers.³ Posting a new article to the timeline had a ritual feel to it, as though one were depositing a prized object into a fast-flowing stream. Likes, replies, and retweets enabled authors to see the reception of their ideas unfolding in real time. Watching them accrue could provide an almost narcotic high for those whose sense of worth was bound up in what they wrote.⁴

Posting’s confusion of word and deed, combined with the vast new reading audiences that the social media economy seemed to open up, gave the intelligentsia the heady sense that its stock and trade—“the discourse”—now offered the route to real and meaningful power.⁵ This professional hubris would result in a series of historical misinterpretations throughout the 2010s, from the Arab Spring to Russiagate, each of which represented an overestimation of the power of language (and social media) to bring about revolutionary political change.

Of course, Twitter had other, rather more dubious pleasures. Writing in 2021, the Scottish academic and publisher Udith Dematagoda called it “a website that seemingly exists for the sole benefit of snitches and hysterics.”⁶ Dematagoda’s targets were the seemingly daily pile-ons, brigading incidents, and cancellation attempts that Twitter was notorious for during this period. His wider point, however, was how undignified and basically infantile all of this activity was. Not only was the language and behaviour of cancellation mobs eerily reminiscent of the primary school playground, but so too were the overall aims of those involved:

Nobody has stopped to consider a very simple premise: that these people are acting like children … Much like children, those most active and successful on twitter merely crave attention and recognition, seemingly lacking the internal resilience required for introspection and ego formation. The result is a dysfunctional hive mind of undisguised ambition in the service of meagre approbation … what is most important is the desultory dopamine high which comes through online validation by one’s peers.

What made the spectacle of 2010s-era cancellation additionally cringeworthy was the fact that many of the most eager participants in online mobs (as well as their victims) were members of the intelligentsia themselves. Encouraged by late 2000s and early 2010s trends such as conference panel live-tweeting and the general need to “have an online profile,” academics and professional groups flocked to Twitter from the early 2010s onwards, creating quasi-professional, platform-based social networks such as academic Twitter, library Twitter, and publishing Twitter. Initially, these were informal groupings where members could promote themselves and take part in friendly discussion. However, despite these groups’ origins in offline communities and professions, the real-time behaviour of their members online nevertheless became governed by the new ethos of social surveillance associated with the digital, as well as the attention economy and notoriety-seeking incentives of the social media stream.

Based on her experiences on message boards during the early 1990s, the internet commentator Carmen Hermosillo realised early on that “digital spaces need divas.” Each board community had its own controversial figures, the “board hos,” who generated the daily drama. The board hos were not at all incidental to proceedings. Rather, they were precisely what kept people coming back to the boards and could therefore have an oddly untouchable status, despite their extreme behaviour. As Katherine Dee puts it, “drama drives posting, and posting is what keeps people online.”⁷ Anyone who becomes central to the daily dramas of group posting can therefore can make themselves indispensable.

The basic logic and inbuilt incentives of social media, with its need to harness people’s time and attention and its corresponding ability to reward time wasters and attention seekers with the adulation they crave, were such that they could awaken playground bully and drama-driven diva tendencies among the members of any social group. Nor indeed was it simply drama that online bullies and divas wanted; combined with drama was a desire for control. The drama tended to be accompanied by new rules, new creeds, and new statements of moral orthodoxy, typically enforced by the systematic targeting and purging of users who failed initially to comply. It was fear of the combined righteous moralism and unpredictable libertinism that the group bullies and divas wielded which generated conformist behaviour among the rest of the group. The new conformism expressed itself both as a denial that any of this was even happening and an assurance that each of the serial ejectees had truly deserved their fates. The intelligentsia proved to be no exception to this rule. Nor, indeed, as Mark Fisher observed in a 2013 essay that in turn got him temporarily cancelled, did the many self-identifying anti-capitalists who likewise flocked to the platform in the early 2010s. They, too, forgot that they were on “enemy territory” in “capitalist cyberspace,” and blithely allowed their own online behaviour to be overwritten with its Janus-faced, simultaneously competitive and hyper-conformist logics.⁸

Social media profiles are typically curated to appeal to the expectations of outside observers. To maintain their reputations online, successful profile holders need to conform to “what is seen to be seen as good,” anticipating the “ever-shifting moral trends” taking place in their communities and shaping their behaviour and language to fit.⁹ Moreover, online reputation is to a large degree determined by a phenomenon I call metadatafication. This describes the way in which one’s status online can be reduced to one’s accompanying descriptions and affinities—one’s affiliation with an institution, for instance, or the impression given by the other users one chooses to follow, or the ideological flavour of the content one chooses to post or retweet. The accompanying social pressures generated by metadatafication have proven especially acute in highly competitive, professionally affiliated online communities such as academic Twitter and publishing Twitter. These are parts of the platform where the online reputation of any given named member is underwritten in a particularly direct way by their offline affiliations and professional status, and where there can be severe consequences for those judged to have made serious discourse mistakes online.

Reputation crises of this sort are neither solely personal nor are they easily contained. Because of the densely interlinked nature of online reputation, a reputational crisis for a newly cancelled academic, for instance, can become a ramifying second-order crisis for anyone in their online and offline networks—for their university, their departmental colleagues, and their students, some or all of whom may feel morally implicated by their suddenly awkward association with the cancelee. The hierarchically mixed nature of these online communities, which can include not only established professional “divas” but also postgraduate students, junior employees, and professional aspirants eager to make their mark and “take down” established figures who prove themselves to be “out of sync” with contemporary moral sentiments, contribute to the intensity and drama of social media reputation spirals, dogpiles, and brigading incidents.

The result of these dynamics by the mid-2010s was the rollout across multiple online communities of what Freddie deBoer called “the planet of cops” mentality. This referred to the seemingly hegemonic culture of mutual suspicion and hypervisibility associated with cancellation events:

a world of snitches, informants, rats … Endless surveillance. Everybody is to be judged. Everyone is under suspicion. Everything you say is to be scoured, picked over, analyzed for any possible offense … digging deeper and deeper to find out who’s Good and who’s Bad.¹⁰

These tendencies would only accelerate. The global lockdowns and workplace furloughs associated with the COVID-19 pandemic drove many more people into social media addiction. Twitter’s own figures showed that “daily active usage” on the platform rose by more than 34% between April and June 2020, a fact that the company attributed to the “increased global conversation around the COVID-19 pandemic and other current events.”¹¹ 2020 brought with it two additional phenomena, the global “racial reckoning” that took place in the wake of George Floyd’s murder and the November 2020 US presidential election. The “global conversation” surrounding these events, combined with the increasingly divisive nature of the lockdown policies and vaccine rollouts themselves, generated an uncanny sense that society was now splitting down the middle on every conceivable new issue.¹²

At more or less the same time, Twitter’s moderation policies became increasingly interventionist and hardline. From about 2018 onwards, gender-critical accounts, including those of a number of academics, were targeted for suspension under Twitter’s “hateful conduct” policies for violations such as misgendering.¹³ In accordance with the logic of metadatafication, the fact that one’s account had been banned from the platform could in turn become a kind of label in itself—a badge of one’s supposedly proven extremism—which could be used to question one’s reputation or standing in other professional contexts. The University of Melbourne philosopher Holly Lawford-Smith found herself being “tagged with an advisory” in this way by Australian journalists after it became known that her Twitter account had been permanently suspended in 2019 for misgendering. There was, she found, a certain innuendo involved, one which extended from the fact of suspension itself to wider assumptions about character and professional suitability. Your account was permanently banned from Twitter for “hateful conduct.” What kind of person are you? Should you really be teaching at a university?¹⁴

After the riots at the Capitol buildings in Washington DC on 6 January 2021, Twitter not only permanently suspended the personal account of Donald Trump “due to the risk of further incitement of violence.”¹⁵ It also suspended those of more than 70,000 of his supporters, ostensibly in order to stop the spread of misinformation and QAnon conspiracies, while advising its own employees “to take care of their mental health.” However, such was the distrust of the platform among those on the right that many suspected that Twitter was simply taking the opportunity presented by the crisis to “crack down on conservatives” in general.¹⁶ It became commonplace in right-wing and “post-left” Twitter circles during this period to screenshot the blank profile pages of newly suspended accounts and post them to the timeline as a kind of memorial and rallying cry: “look what they did to my boy.”

Perhaps because their own accounts and follow circles were almost never affected by these penalties, the progressive intelligentsia underestimated the long-simmering resentment and sense of unequal treatment that Twitter’s content moderation policies generated among those who were not part of the progressive orthodoxy. They were also, perhaps, unprepared for how Elon Musk would seek to leverage this resentment to curry favour with Twitter’s former outgroups when he bought the platform in 2022. Musk’s attempt to shift the new “X” away from what he represented as the old Twitter regime’s “extreme left” orientation extended most obviously to boosting content produced by right-wing conservatives and members of the J. D. Vance-aligned “new right.” This in turn affected the composition of the X user base: according to the 2025 Reuters Digital News Report, the number of American right-wingers who used X for news had risen to 26% by early 2025, as opposed to only 9% under the previous ownership regime in 2021.¹⁷

However, the company’s decision (in November 2022) to offer an amnesty to suspended account holders also brought gender-critical feminists, former “dirtbag leftists,” and an array of lockdown and vaccine sceptics of assorted ideological pedigrees back into the fold.¹⁸ Lawford-Smith, for instance, regained her Twitter account in late 2022, as did other formerly “high volume” and politically heterodox users whose accounts had been suspended under the old content moderation policies, including the comedian Graham Linehan, the podcasters Meghan Murphy and Aimee Terese, the philosopher Jordan Peterson, and the celebrity author and contrarian Naomi Wolf. Similarly, although the Reuters Digital News Report showed that the number of left-wing American users who visited X at least once a week for news had collapsed to a low of 15% in 2024 (as compared to 28% in 2020), it had risen again to 24% by early 2025.¹⁹

Those within the intelligentsia and legacy media who had constituted the old Twitter in-group, however, viewed Musk’s statements and policy changes with distaste and a certain slowly escalating horror. Reporting on the new amnesty policy, The Guardian quoted Imran Ahmed, chief executive of The Center for Countering Digital Hate, an NGO with a long history of petitioning social media companies to suspend the accounts of controversial public figures on hate and misinformation grounds.²⁰ “Superspreaders of hate, abuse and harassment will be the only people to benefit from this latest decision by Twitter,” The Guardian quoted Ahmed as saying. “The choice for advertisers has never been starker: stick around and back Elon Musk, or protect their brands and ensure their marketing dollars aren’t used to enable the spread of hate, abuse and disinformation.”²¹

Over the following two and a half years, a number of news companies, charities and NGOs, academic institutions, corporate accounts, and arts organisations either closed their accounts or stopped posting to them, often issuing public statements as they did so.²² The rhetoric used in these statements is in itself revealing. When The Guardian mothballed its editorial X accounts shortly after the US presidential election in November 2024, it justified its decision on the basis of “the often disturbing content promoted or found on the platform, including far-right conspiracy theories and racism.” However, the result of the election (and Musk’s then role as a high profile supporter of Trump) appeared to have precipitated the accounts’ closure:

The US presidential election campaign served only to underline what we have considered for a long time: that X is a toxic media platform and that its owner, Elon Musk, has been able to use its influence to shape political discourse.²³

Other leaving statements likewise highlighted “problematic” content on the platform (and the figure of Musk himself), but stated more explicitly that X was now out of alignment with their corporate values. The British domestic violence charity Respect claimed that its decision to leave X in January 2025 was motivated by “growing concerns about the use of the platform, both by its users and its owner, to cause and encourage harm, promote misogyny and spread disinformation.” However, its statement also reiterated the charity’s wider moral value system: “Our vision is a world where it is never ok to control, harm or cause fear” and “We prioritise the safety and wellbeing of our staff, members and the wider community.” It was the clash between these values and the platform as it now was that meant that “We can’t, in good conscience, continue to engage on X in its current form.”²⁴ For the School for Social Entrepreneurs—which self-identifies as “inclusive, supportive, and purposeful”—the new “climate” on X “no longer aligns with our vision. Harmful rhetoric, divisive content and misinformation have become commonplace” and “this shift goes against our fundamental values.” Accordingly, X was “no longer a platform that we want to be part of.”²⁵ PEN Scotland, meanwhile, abandoned its X account on the basis that “some writers and readers would be more comfortable following us elsewhere.”

The point of mounting a consumer and advertising boycott campaign against X, as was periodically attempted between 2022 and 2025, was to reverse the network effects that had drawn people to Twitter in the first place. Rather than being somewhere that users had to have a presence on if they wanted to connect with the “big names” and keep up with “the discourse,” X would, as Helen Lewis puts it, now “mark the spot” where “large numbers of people no longer wanted to be.”²⁶ It is, of course, undeniable that much of what the leavers had to say about the kind of content now being posted and boosted on X is simply true. As Lewis jokingly (but accurately) observed in July 2023, “scrolling Twitter’s ‘For You’ tab has become like bobbing for apples in a bowl full of amateur race scientists and Roman-statue avatars lamenting that we no longer build cathedrals.”²⁷ Anyone who has remained on the platform can confirm that the experience has not improved in that regard in the two years since. However, despite sustained attempts to render X, in the words of Imran Ahmed and The Guardian, “toxic,” the platform itself has proved hard to kill. Recent data indicates that its user base has stayed relatively stable despite the boycott campaigns and the arguable decline of the platform user experience. 11% of global users surveyed in 2025 report using it for news at least once a week, as against 2% for Bluesky.²⁸

Even as former account holders fled the platform, they bore the linguistic and logical imprints of Twitter with them, testifying to its staying power. Their leaving statements were linguistically compressed and emotion-laden statements of institutional positionality, reflecting how social media has taught both individuals and organisations to “speak.” The underlying logic of these statements is that of metadatafication: affiliation and alignment, the link, and the descriptive tag. Who do we stand with (or more precisely, perhaps, who do we not stand with)? What do the relationships we make visible to others say about our reputation and who we are? Anxieties about “bad connections” slip into the poorly defined, but affectively heightened, language of social harm.²⁹ All of this reflects what Byung-Chul Han calls the overwhelming “lateral anxiety” produced by contemporary culture, whereby people assess their value through “constant comparison with others.”³⁰ What Twitter was, and what social media continues to be, is a redescription of the world in terms of networks and relationships, and a sense imparted on its users that we lack validity and authenticity unless we are plugged securely into the right channels.

1. Helen Lewis, “The Weird, Fragmented World of Social Media After Twitter,” The Atlantic (30 July 2023).

2. Karl Mannheim, Ideology and Utopia: An Introduction to the Sociology of Knowledge, translated by Louis Wirth and Edward Shils (1936; San Diego: Harcourt Brace, 1985), 10.

3. Ryan Ruby, “A Golden Age?”, Vinduet (24 April 2023).

4. See Rob Horning, “Social Media, Social Factory,” The New Inquiry (29 July 2011); and Marion Fourcade and Kieran Healy, The Ordinal Society (Cambridge, MA: Harvard University Press, 2024), 63.

5. Cf. Bernice Martin, A Sociology of Contemporary Cultural Change (Oxford: Basil Blackwell, 1981), 20-1.

6. Udith Dematagoda, “On Infantalism: Reflections on Art, Morality, and Regressed Life,” Immanent Dissolution (5 March 2021).

7. Katherine Dee, “Selling the Drama,” The New Atlantis, no. 72 (2023): 93.

8. Mark Fisher, “Exiting the Vampire Castle,” in K-Punk: The Collected and Unpublished Writings of Mark Fisher, edited by Darren Ambrose (London: Repeater Books, 2018), 744.

9. Hans-Georg Moeller and Paul J. D’Ambrosio, You and Your Profile: Identity After Authenticity (New York: Columbia University Press, 2021), 28.

10. Freddie deBoer, “Planet of Cops,” Freddie deBoer (26 August 2021). This essay was originally written in May 2017.

11. Twitter, Q2 2020 Letter to Shareholders, 23 July 2020.

12. Lewis, “The Weird, Fragmented World of Social Media After Twitter.”

13. Holly Lawford-Smith, Sex Matters: Essays in Gender-Critical Philosophy (Oxford: Oxford University Press, 2023), 175.

14. Holly Lawford-Smith, Gender-Critical Feminism (Oxford: Oxford University Press, 2022), 122–3.

15. “The 65 Days That Led to Chaos at the Capitol,” BBC News (10 January 2021).

16. Tony Romm and Elizabeth Dwoskin, “Twitter purged more than 70,000 accounts affiliated with QAnon following Capitol riot,” Washington Post (12 January 2021).

17. Dan Williams, “The Decline of Legacy Media, Rise of Vodcasters, and X’s Staying Power,” Conspicuous Cognition (25 June 2025).

18. Dan Milmo, “Elon Musk offers general amnesty to suspended Twitter accounts,” The Guardian (24 November 2022).

19. Williams, “The Decline of Legacy Media.”

20. For the CCDH’s prior role in petitioning Twitter to suspend particular user accounts, see, for instance, Ash Percival, “Katie Hopkins’ Twitter Reinstated Following Week-Long Absence,” Huffington Post (2 July 2020); and “Twitter bans David Icke over Covid misinformation,” BBC News (4 November 2020).

21. Milmo, “Elon Musk offers general amnesty to suspended Twitter accounts.”

22. The rationale for leaving an institutional account dormant rather than deleting it outright is to prevent subsequent account impersonation by copycats or opportunists. This seems to have become “best practice” among those leaving the X platform.

23. “Why the Guardian is no longer posting on X,” The Guardian (13 November 2024).

24. “Why Respect is no longer posting on X,” Respect (28 January 2025).

25. David McGlashan, “Why we’re no longer posting on X/Twitter,” School of Social Entrepreneurs (15 November 2024).

26. Lewis, “The Weird, Fragmented World of Social Media After Twitter.”

27. Lewis, “The Weird, Fragmented World of Social Media After Twitter.”

28. Williams, “The Decline of Legacy Media.”

29. See Thomas Raymen, The Enigma of Social Harm: The Problem of Liberalism (London: Routledge, 2023).

30. Byung-Chul Han, The Expulsion of the Other, translated by Wieland Hoban (Cambridge: Polity Press, 2018), 32.