The Rage Machine: How Platforms Converted Moral Emotion into a System of Visibility, Identity and Mobilisation
Culture & Society · Episode 04 · August 2026
The Rage Machine
How platforms converted moral emotion into a system of visibility, identity and mobilisation — and what that has done to public life
Nelly Bekono · NellyBekono.com · Echoes of Changes — Culture in Flux
There is a familiar way of speaking about social media that begins with individual weakness. People are impulsive. They enjoy conflict. They believe what flatters them. They have lost the capacity to concentrate. Each statement contains a measure of truth, yet together they produce a remarkably convenient explanation. If the digital public sphere is degrading, the fault belongs to its users. The architecture disappears from view. That is precisely where a serious analysis must begin.
Human beings brought old emotions into a new environment, but the environment did not receive those emotions neutrally. It sorted them. It assigned them differential value. It learned that some expressions generate faster reactions than others, that certain conflicts keep people present, and that moral accusation is exceptionally effective at turning spectators into participants. What looks from the outside like a spontaneous eruption is often the meeting point of human feeling and automated selection.
This essay calls that meeting point the rage machine. The phrase does not describe a secret control room, a single algorithm or a conspiracy by technology companies to make everyone furious. It describes a system of incentives. Platforms compete for attention; ranking systems predict what will hold it; users learn which performances travel; political actors, advertisers and creators adapt; the resulting behaviour becomes new data for the system. No participant controls the whole arrangement, but every participant is conditioned by it.
The problem, then, is not that outrage exists. A society incapable of outrage would be morally anaesthetised. The problem is that digital systems have made outrage unusually available, legible and profitable — and have done so without developing comparable incentives for memory, context, proportionality or repair.
In 1969, ARPANET connected computers at four American research institutions, including the University of California, Los Angeles and the Stanford Research Institute. On 29 October of that year, a UCLA team attempted to transmit the word “LOGIN” to a computer at Stanford; the system crashed after the first two letters, which meant that “LO” became the first message sent across the network. The event was modest, but the principle behind it was decisive: information could move between distant computers through a shared technical network rather than remaining confined to one machine or institution.
During the 1970s, that infrastructure began acquiring forms of communication that ordinary users would eventually recognise. Networked email expanded after Ray Tomlinson adapted electronic messaging for ARPANET in 1971 and used the @ symbol to separate the user from the host computer, while bulletin-board systems allowed people with a modem and a telephone line to leave messages and files for others. Usenet, created in 1979 by Tom Truscott and Jim Ellis at Duke University, organised discussion into subject-based newsgroups and demonstrated that a network could support communities built around continuing conversation rather than geographical proximity.
These early systems were limited, technically demanding and available to a small population, yet they established practices that would later define digital social life: a persistent identity, a public message, a reply, a group organised around shared interests and an archive through which the conversation could be revisited. The technology was not yet mass social media, but the underlying social discovery was already visible, because people learned that a network designed to exchange information could also become a place in which relationships, reputations and communities developed.
The next transformation came from Geneva. In March 1989, while working at CERN, Tim Berners-Lee proposed a system that would connect information through hypertext, and by the end of 1990 he had built the first web server and browser on a NeXT computer. The first website, hosted at info.cern.ch, explained the World Wide Web project itself, while CERN’s decision on 30 April 1993 to place the Web software in the public domain allowed others to use and develop it without paying a licensing fee. Later that same year, the Mosaic browser made the Web far easier to navigate on personal computers, helping move it beyond scientific institutions and towards a mass public.
The culture of the early Web during the 1990s was markedly different from the platform environment that would follow. Users often had to search deliberately for websites, join a particular forum or construct a personal page, which meant that online activity was distributed across many independently operated spaces. GeoCities, founded in 1994, allowed users to build home pages within themed digital “neighbourhoods”; services such as ICQ, launched in 1996, made real-time messaging more accessible; and blogging tools at the end of the decade gave individuals an easier way to publish continuously without building every page by hand.
The first recognisable social-networking site arrived in 1997, when SixDegrees combined personal profiles, visible lists of friends and the ability to move through other people’s networks. Those features now appear ordinary, but bringing them together changed the organising principle of online social life: the network itself became visible. Instead of meeting only through a topic or forum, users could display who they knew, discover connections through other people and present a stable version of themselves within the same system.
SixDegrees closed in 2000, partly because the wider internet population and the technology supporting continuous participation had not yet reached the scale required for such a service to become indispensable. Nevertheless, the model survived. Friendster, founded by Jonathan Abrams in 2002 and launched publicly in 2003, grew rapidly by allowing users to connect through friends, although technical difficulties made that growth difficult to sustain. LinkedIn launched in 2003 with a professional emphasis, while MySpace, also founded in 2003, combined profiles, music, personal customisation and visible friendship networks in a way that became especially attractive to young users and emerging artists.
Facebook entered this already developing field on 4 February 2004, when Mark Zuckerberg, Eduardo Saverin, Dustin Moskovitz and Chris Hughes launched TheFacebook at Harvard University. It initially expanded through universities, tied identity to institutional email addresses and presented itself as a network of real social connections rather than an anonymous public forum. By December 2004, Facebook reported one million active users, and in September 2006 it opened registration beyond educational institutions to anyone aged thirteen or older with a valid email address.
Yet the importance of Facebook did not lie only in the number of people who joined. On 5 September 2006, the company introduced News Feed, which brought updates from friends onto the user’s home page instead of requiring people to visit each profile individually. This altered the rhythm of the social Web because the platform now assembled other people’s activities into a continuous stream, while the user moved from deliberately seeking a person or page towards receiving an organised account of what was happening across the network.
Other platforms were transforming the form of public expression at the same time. YouTube registered its domain in February 2005, and on 23 April co-founder Jawed Karim uploaded “Me at the zoo,” a nineteen-second video recorded at the San Diego Zoo. The technical achievement was not simply that video appeared online, since online video already existed, but that uploading and distributing it became accessible to people without a broadcaster, studio or personal server. In March 2006, Jack Dorsey sent the first message on a service then called “twttr,” and Twitter subsequently developed a public, real-time conversation organised around short posts, replies, hashtags and trending subjects.
By the middle of the 2000s, the architecture of contemporary social media was therefore taking shape through several different technologies: Facebook organised existing relationships through the social graph; YouTube made user-generated video globally distributable; Twitter made short public reaction immediate; and blogs continued to support longer individual publication. People were attracted because these services answered real needs, allowing families to remain connected across distance, marginalised communities to find one another, unknown creators to reach audiences and citizens to speak publicly without first being selected by an editor or broadcaster.
The transformation accelerated in 2007 when Apple introduced the iPhone as a mobile phone, an iPod and an internet communications device within the same touchscreen product. It went on sale in the United States on 29 June 2007, and the App Store followed in July 2008, allowing third-party applications to be installed and updated through one distribution system. Smartphones did not create social media, but they removed the practical separation between being online and moving through everyday life, because a camera, internet connection, communication device and growing collection of platform applications could now remain with the user throughout the day.
Facebook released its iPhone application in July 2008 and introduced the Like button in February 2009, turning a broad range of reactions into one frictionless and countable signal. Instagram, created by Kevin Systrom and Mike Krieger, launched in October 2010 around mobile photography, while Snapchat followed in 2011 with disappearing images that responded to growing discomfort with the permanence of online self-presentation. Facebook announced its acquisition of Instagram in April 2012, and by October of that year it reported more than one billion monthly active users.
The speed of adoption is visible in survey data. Pew Research Center found that only 7 percent of American adults used a social-networking site in 2005, whereas 65 percent did so by 2015; among adults aged eighteen to twenty-nine, adoption increased from 12 percent to 90 percent during the same decade. Internet use itself had grown from 52 percent of American adults in 2000 to 84 percent in 2015. These are United States figures rather than global measurements, but they show how rapidly an activity associated with a relatively small population became part of ordinary social life.
As the number of users and publications expanded, chronological presentation became increasingly difficult to sustain. Facebook’s News Feed had used ranking for years, and other platforms gradually moved towards systems that predicted which content each user was most likely to value or engage with. Instagram explained that by 2016 users were missing around 70 percent of the posts in their chronological feeds, which it used to justify introducing ranking. The practical problem was real: no person could process everything produced by a large network. Yet the solution gave platforms a new form of authority, because they no longer merely hosted available material but continuously decided which part of that material would become visible to each individual.
The data required for those decisions came from behaviour. A pause on a video, a click, a search, a Like, a comment, a share or the decision to follow an account could become a signal from which the system inferred preference. The recommendation then influenced what the user encountered next, and that new encounter generated further behaviour from which another prediction could be made. The result was not a static portrait of what a person already liked but a feedback process in which the platform’s selection and the user’s response continually shaped one another.
This predictive capacity was connected to an advertising model that had become central to the largest services. Facebook launched its self-service advertising platform in November 2007, allowing advertisers to target audiences through information associated with profiles and behaviour. As platforms accumulated larger quantities of data, advertising could be directed towards increasingly specific groups, while engagement supplied both more time in which to show advertisements and more information with which to refine targeting. In 2024, the United States Federal Trade Commission concluded after examining major social-media and video services that targeted advertising accounted for most of the revenue of many of the companies it studied and created incentives for continuous data collection and content designed to increase user engagement.
The shift from a social graph towards recommendation became even clearer with short-form video. ByteDance launched Douyin in China in 2016 and introduced TikTok internationally in 2017, later combining it with Musical.ly in 2018. Unlike earlier networks that primarily began with a list of friends or accounts chosen by the user, TikTok’s For You feed could begin recommending videos immediately and refine the selection through signals such as viewing behaviour and interaction. The central object was no longer the person one had decided to follow but the item most likely to retain attention at that moment.
This technological development changed the conditions of visibility without requiring any engineer to decide that society should become more emotional or divided. Systems optimised for response learned which forms of material produced measurable activity, while research found that moral-emotional language and hostility towards an opposing group circulated especially effectively in political communication. Anger demanded an answer, fear created urgency and moral accusation invited people to take a side, whereas a careful explanation required time and the willingness to remain uncertain before responding.
People could observe the consequences directly. Creators compared the reach of different videos, journalists saw which headlines travelled, political actors identified which statements dominated discussion and ordinary users noticed which versions of their experiences received recognition. The metrics beneath each publication turned communication into a visible experiment, so a person did not need to understand the recommendation system in order to adapt to its results. Over time, the expressions favoured by the platform could appear to represent what the public wanted, even though their prominence had also been produced by the system selecting and distributing them.
Nevertheless, people continued to participate because the platforms still delivered genuine value. They sustained relationships across distance, enabled professional and creative opportunities, gave communities a place to organise and allowed people excluded by traditional gatekeepers to speak publicly. The difficulty is therefore not that a harmful technology replaced an entirely healthy social world, but that valuable forms of connection developed inside infrastructures whose commercial incentives increasingly depended on prediction, measurement and the continuous retention of attention.
By the beginning of the 2020s, social media was no longer one clearly bounded category of website. Messaging, shopping, entertainment, news, political communication, professional networking and audiovisual search increasingly overlapped within the same applications, while established institutions adapted their work to platform distribution. News organisations reformatted reporting for feeds, universities compressed knowledge into short videos, companies treated engagement as evidence of relevance and political leaders communicated directly with publics whose reactions could be measured in real time.
This is how social media moved from being a collection of tools into an environment that participates in organising social life. The path ran from ARPANET’s first two-letter message in 1969 to the open Web of 1993, from SixDegrees’ visible friendship network in 1997 to Facebook’s News Feed in 2006, from the iPhone in 2007 to the Like button in 2009, and from the social graph to recommendation-driven feeds during the following decade. At every stage, a technical solution answered a genuine need, yet each solution also changed what could be measured, predicted and made visible.
Only after following that development can we properly ask what happened once platforms became primary mediators of emotion and identity, because outrage, micro-tribal belonging and the public performance of the self did not arrive from outside the history of digital connection. They emerged from the interaction between human desires for recognition and belonging, technologies that made social response measurable, and business models that assigned economic value to the ability to hold attention.
1. Before the Feed: Outrage Has a History
Why technological determinism misses the point
Every generation is tempted to mistake the technologies of its own period for the origin of the behaviours those technologies reveal. Outrage existed long before a person could express it with a tap. Religious denunciations, revolutionary pamphlets, partisan newspapers, radio demagogues and televised moral panics all converted grievance into attention and collective action. The nineteenth-century press understood that scandal sold papers. Twentieth-century broadcasters learned that conflict concentrated audiences.
This history matters because it prevents technological determinism. Social media did not introduce propaganda, faction, hatred, rumour or public shaming. Nor did it eliminate the gatekeepers that preceded it so much as replace a relatively visible set of editorial gates with a largely invisible and continuously adjusted system of ranking. The newspaper editor decided what entered tomorrow’s edition. The feed makes millions of such decisions for millions of people, moment by moment, using signals gathered from their own behaviour.
The decisive shift was gradual. As platforms expanded, the quantity of available material exceeded what any user could process. Ranking became necessary — but ranking also became commercial. The question was no longer simply what had been posted most recently. It was what a particular person was most likely to click, watch, share or answer.
“Here lies the first structural advantage of outrage: it produces behaviour that machines can easily read. Nuance often appears as hesitation. Outrage produces visible action.”
The digital transformation of outrage is best understood as a change in availability. Anger no longer has to wait for a national scandal or the publication schedule of a newspaper. A user can enter an emotionally charged public arena before getting out of bed. The question is not whether the appetite is ancient. It is how an environment identifies, stimulates and monetises it.
2. Why Outrage Wins the Contest for Attention
Speed, visibility, identity, mobilisation
Not every emotional message becomes visible. Yet research consistently indicates that high-arousal emotion has a distributional advantage over quieter states. Jonah Berger and Katherine Milkman’s study of highly shared New York Times articles found that arousal helped explain virality: material producing awe, anxiety and anger was more likely to be shared, while sadness was associated with lower sharing. The important distinction was not positive versus negative, it was activation versus stillness.
William Brady and colleagues later examined more than half a million tweets on gun control, same-sex marriage and climate change. They found that each additional moral-emotional word was associated with approximately twenty per cent greater diffusion — and that transmission was especially pronounced within ideological networks. Outrage did not merely spread information. It travelled along existing lines of affiliation.
Speed. Outrage offers a rapid interpretation of an event. It identifies a violation, assigns responsibility and supplies an emotional posture before uncertainty has time to become uncomfortable. A complicated episode becomes a simple moral shape: victim, offender, witness, duty.
Visibility. Outrage is public-facing. Its performance involves a share, denunciation, hashtag or visible alignment. Careful uncertainty may result in no platform action at all. The ranking system does not reward epistemic care because much of that care remains invisible.
Identity. A moral reaction communicates more than an opinion. It announces what kind of person the speaker is, which community the speaker recognises, and which violations that community refuses to tolerate. Outrage is a compressed identity statement: I know what matters; I recognise the enemy; I can be trusted to react correctly.
Mobilisation. Outrage lowers the cost of joining collective action. One does not need membership in an organisation, attendance at a meeting or mastery of a long argument. A visible reaction is enough to enter the crowd.
“This is conditioning without command. The platform need not instruct users to become angrier. It need only distribute rewards unevenly. The sharp formulation travels; the qualified one stalls.”
A personalised feed creates a private front page whose selection criteria the reader cannot fully inspect. This produces an illusion of prevalence: if several emotionally similar items appear in succession, the issue can feel omnipresent even when assembled from a narrow slice of available material. The feed does not only show what is happening. It changes the perceived scale and emotional temperature of what is happening.
3. From Opinion to Performance
When identity hardens into brand
Public identity has always involved performance. Social media added a continuous public ledger. The profile accumulates declarations, affiliations, images, reactions and silences. A political or moral identity is no longer expressed only in decisive moments; it is maintained through an ongoing sequence of visible choices.
In this environment, outrage functions as proof of membership. React too slowly and one may appear indifferent. Introduce context too early and one may appear disloyal. The reaction becomes an examination administered in public.
Steve Rathje, Jay Van Bavel and Sander van der Linden analysed posts by political figures and news organisations and found that language about the political out-group was a particularly strong predictor of sharing and engagement. Negative partisanship — the attachment produced by hostility to an opposing side — can be easier to activate than agreement on a common programme.
The difficulty arises when platform incentives favour the most legible, emphatic and conflict-oriented version of identity. Complex commitments become branding. Identity hardens precisely when democratic life requires it to remain revisable.
Gamergate remains an instructive case. Beginning in 2014, arguments under the language of “ethics in games journalism” intertwined with sustained harassment directed at women in game development, including Zoë Quinn, Anita Sarkeesian and Brianna Wu. It showed how a networked conflict can train participants, generate shared enemies and migrate into broader political identities. Their coherence is often affective before it is ideological.
“Outrage resolves the problem of recognition. In a vast and anonymous network, how do strangers know who belongs? They recognise a shared object of condemnation. Moral emotion becomes a social passport.”
4. Micro-Tribalisation: Belonging After the Institution
How identity clusters form around sensitivities, not histories
Micro-tribalisation describes the formation of small or narrowly coded identity clusters organised around shared sensitivities, symbolic fluency and repeated emotional alignment. Its defining feature is not simply that members share information. It is that belonging is continuously tested through reactions to signals.
Earlier forms of belonging possessed institutional thickness: durable roles, face-to-face obligations, procedures for decision, memories extending beyond the latest dispute. Digital micro-tribes offer belonging without bureaucracy, identity without long apprenticeship and collective feeling without a shared life.
This lightness is part of their appeal. A person can find an intelligible social world in minutes. A meme condenses history into recognition. A hashtag gathers dispersed experience. A phrase becomes a password. For someone denied recognition elsewhere, this can be genuinely liberating.
Yet compressed belonging carries a hidden cost. When the group lacks durable institutions, emotional alignment must do more work. The result can be escalating purity. The almost-insider — the person who knows the language but questions the group’s interpretation — may provoke more anger than the declared opponent.
Social psychology calls this the black sheep effect: groups judge deviant members of their own category more harshly than comparable outsiders because internal deviance threatens the group’s positive identity. Online, that judgment unfolds before an audience and can be quantified through likes, reposts and replies.
The 2024 conflict between Kendrick Lamar and Drake illustrated this at scale. Diss tracks were dissected as dossiers, memes served as allegiance markers, and judgments about artistry merged with claims about character and morality. Cultural appreciation became a referendum on personal loyalty.
“Micro-tribalisation is not caused by algorithms alone. Loneliness, institutional distrust and the weakening of intermediary institutions all create demand for belonging. Platforms meet that demand with an abundant but unstable supply.”
5. Mobilisation Without Organisation
When the crowd forms before the structure
Outrage has always mobilised, but platforms dramatically reduce the cost and time required to gather a crowd. The #MeToo movement connected testimony that institutions had treated as isolated, making patterns of abuse publicly visible. Black Lives Matter used networked communication to document violence and organise protest. Any analysis that treats mobilisation itself as pathological would reproduce the perspective of institutions that prefer publics to remain quiet.
The question is not whether outrage mobilises but what happens when mobilisation becomes easier than organisation. Organisation requires decisions about goals, roles, evidence, tactics, accountability and duration. Networked mobilisation can gather an enormous public before those questions are answered. The crowd forms first; structure, if it comes at all, follows.
A rapidly assembled public may exert necessary pressure, but it can also misidentify a person, spread an edited clip or substitute punishment for reform. Participation becomes detached from responsibility. Each user contributes only a small act, while the aggregate effect can include threats, loss of employment or physical danger.
The attack on the United States Capitol on 6 January 2021 is the most extreme case. A 2024 study by Huidong Li and colleagues found that specific communications predicted bursts in the level and duration of violence and in rioters’ use of weapons — showing how digital messages can operate as coordination signals inside an already mobilised crowd.
“Political entrepreneurs use ambiguity strategically — issuing language strong enough to activate supporters while retaining deniability about specific consequences.”
6. Falsehood as a Social Instrument
Why corrections so often fail
The conventional language of misinformation treats falsehood as a defect in the information supply. That model is useful but incomplete. Information also performs social work. A story can identify allies, explain misfortune, intensify distrust, protect a valued identity or demonstrate loyalty. Its factual weakness may be less important to its circulation than its social usefulness.
Soroush Vosoughi, Deb Roy and Sinan Aral’s influential 2018 study examined approximately 126,000 cascades of contested stories on Twitter. False news travelled farther, faster, deeper and more broadly than true news — particularly in politics. The difference in spread was driven substantially by human behaviour, not bots.
Outrage gives falsehood a motor. A claim framed as a violation asks the user to react before checking it. The original allegation is simple, vivid and morally complete. The correction is procedural: the photograph came from another country; the document does not say what the caption claims. The false story provides a villain and a feeling. The correction provides metadata.
In Myanmar, Facebook became for many users nearly synonymous with the internet. The United Nations’ independent fact-finding mission documented the role of social media in an environment of incitement directed at the Rohingya. The platforms are global; the feelings they carry are local. A moderation system trained primarily in English may miss coded incitement in a regional language.
“False stories persist not only because users lack facts. They persist because they can stabilise a social world — telling people why events feel confusing, who is responsible and whom to trust.”
7. When Attention Becomes Infrastructure
The business model of moral urgency
Attention has always been economically and politically valuable. What changed was the granularity, continuity and intimacy with which it could be measured. A platform can observe whether a user stopped, clicked, enlarged, replayed, commented, returned or abandoned — and compare those actions across vast populations. Attention ceases to be only an audience outcome. It becomes an operational input.
Calling attention infrastructure means recognising that it now shapes the conditions under which other social activities occur. Roads organise movement without determining every destination. Attention systems organise visibility: they influence which events become perceptible, how long they remain present and what kinds of expression are rewarded.
The feed collapses distinct temporalities into a continuous present. A war, celebrity dispute, personal anniversary and decade-old video can appear within seconds of one another. Context becomes expensive because it requires leaving the stream. Digital culture remembers everything in storage and very little in sequence.
Outrage is especially well suited to this infrastructure because it renews itself. Satisfaction can end a session; indignation points to further evidence, further offenders and further duties. Leaving the platform can feel like abandoning a cause.
“It would be too simple to say that companies sell anger. They sell access to audiences. But if anger reliably contributes to retention, the commercial system has weak internal reasons to suppress it.”
8. The World the Machine Does Not Explain
On keeping structural analysis honest
A compelling theory is dangerous when it explains too much. If every protest or passionate statement is attributed to “the algorithm,” analysis becomes another form of evasion. People do not oppose one another only because a feed made them emotional. Sometimes they have incompatible goals. Sometimes institutions are corrupt. Sometimes a moral violation is clear and delay benefits the powerful.
The rage machine is a theory of distribution and conditioning — not a complete theory of contemporary politics. It does not tell us whether the underlying claim is true, whether a movement is just or whether anger is proportionate. Those remain political and moral judgments.
Outrage about genocide is not equivalent to outrage at the exposure of genocide. A platform may reward conflict symmetrically while the world that produced the conflict is radically unequal. At the same time, a just cause can circulate a false claim. A marginalised community can punish internal dissent. The morality of an objective and the ethics of a method must be evaluated separately.
“Outrage identifies a boundary: this should not be tolerated. Judgment asks what happened, who is responsible, how certain we are, what response is proportionate. The first can initiate moral attention. It cannot complete the work.”
9. Beyond the Ambient Emergency
What a healthier public sphere would require
The deepest effect of the rage machine may be its transformation of atmosphere. Outrage no longer waits for a scandal because the feed keeps a portfolio of unresolved violations continuously available. The user moves through a permanent moral emergency in which attention is both evidence of goodness and a resource others compete to capture.
This atmosphere changes the self. People begin to anticipate an audience while forming a thought. Language is sharpened for circulation. Ambivalence is edited out. It changes relationships, institutions and knowledge. Claims are increasingly sorted by loyalty before they are tested as evidence.
The task is to recover distinctions that the feed continually collapses: between attention and importance, virality and representativeness, accusation and evidence, solidarity and conformity, mobilisation and organisation, moral clarity and moral certainty.
Most importantly, societies must rebuild forms of belonging that do not depend on permanent antagonism. The appetite for micro-tribal belonging is partly a response to deficits elsewhere. If democratic institutions want citizens to choose slower, thicker forms of association, they must make those forms credible and accessible.
“The rage machine meets real needs through a system that extracts value from their intensification. It offers belonging, but often in exchange for an enemy. It offers clarity, but often by removing context.”
Outrage will remain part of public life because injustice will remain part of public life. The goal cannot be emotional neutrality. It must be to prevent a legitimate moral capacity from becoming the default operating system of attention. Anger should alert us to a boundary, not relieve us of the obligation to understand what lies on either side of it.
Somewhere in the last hour, a portion of your attention may have been rented out without a lease you remember signing. The first act of resistance is not silence. It is to become more exact about what we are doing when we look, react and share. The second is to demand environments in which exactness has a chance to survive.
A Note on Method
This essay draws on research in communication, social psychology, political behaviour and computational social science. Its central term, micro-tribalisation, is used as an editorial framework rather than as an established scientific category. The examples are intended to clarify mechanisms, not to imply that different events are morally equivalent. Where studies establish association or prediction rather than causation, the distinction has been preserved. This analysis accompanies Episode 04 of Echoes of Changes — Culture in Flux: “Social Media and Algorithmic Outrage.”
Selected References
Berger, J., & Milkman, K. L. (2012). “What Makes Online Content Viral?” Journal of Marketing Research, 49(2), 192–205.
boyd, d. m. and Ellison, N. B. (2007). “Social Network Sites: Definition, History, and Scholarship.” Journal of Computer-Mediated Communication, 13(1), 210–230.
Brady, W. J., et al. (2017). “Emotion shapes the diffusion of moralized content in social networks.” PNAS, 114(28), 7313–7318.
CERN. “A short history of the Web” and “The birth of the Web.”
Crockett, M. J. (2017). “Moral outrage in the digital age.” Nature Human Behaviour, 1, 769–771.
Fan, R., et al. (2014). “Anger Is More Influential than Joy: Sentiment Correlation in Weibo.” PLOS ONE, 9(10), e110184.
Federal Trade Commission (2024). A Look Behind the Screens: Examining the Data Practices of Social Media and Video Streaming Services.
Li, H., et al. (2024). “Quantifying social media predictors of violence during the 6 January US Capitol insurrection.” Journal of the Royal Society Interface, 21(220).
Meta. “Company Information: Our History.”
Papacharissi, Z. (2015). Affective Publics: Sentiment, Technology, and Politics. Oxford University Press.
Pew Research Center. “Social Media Usage: 2005–2015” and “Americans’ Internet Access: 2000–2015.”
Rathje, S., Van Bavel, J. J., & van der Linden, S. (2021). “Out-group animosity drives engagement on social media.” PNAS, 118(26).
UN Human Rights Council. (2018). Report of the Independent International Fact-Finding Mission on Myanmar.
Vosoughi, S., Roy, D., & Aral, S. (2018). “The spread of true and false news online.” Science, 359(6380), 1146–1151.
YouTube. “Celebrating 10 Years of YouTube.”
Apple. “Apple Reinvents the Phone with iPhone,” 9 January 2007.
Echoes of Changes — Culture in Flux · Episode 04 · NellyBekono.com