Events that moved the world

The Industrial Revolution

Britain from the 1760s; Europe, North America and Japan in the 19th century

Steam power, the factory and the railway moved most people in industrialising countries from farm work to wage work within a few generations, and from villages to cities. Child labour, sixteen-hour days and urban squalor were the first results. Engels's The Condition of the Working Class in England (1845), based on Manchester, is in part an early urban sociology.

How it was read

It was the event that produced sociology itself (see Origins). Its most penetrating later interpreter was the economic historian Karl Polanyi, whose The Great Transformation (1944) argued that the attempt to create a self-regulating market society treated labour, land and money as commodities, when none of them were produced for sale. He called them fictitious commodities. Society reacted to protect itself through factory laws, unions, social insurance and tariffs — a double movement of marketisation and protection.

What lasted

The Luddites, the English textile workers who broke machines between 1811 and 1816, were not opposed to technology as such; historians, notably E. P. Thompson, have shown they were opposed to its use to undercut wages and skilled work. The name has been an insult ever since, often unfairly.

Why it matters now Polanyi is having a revival. Several writers argue that personal data and human creative work are being turned into fictitious commodities by AI, and that the regulation now emerging is the start of a new protective counter-movement.

The French Revolution

France, 1789–99

The fall of the Bastille, the abolition of feudal privileges, the Declaration of the Rights of Man, the execution of the king, the Terror and finally Napoleon. It created the modern vocabulary of politics: left and right (from where deputies sat), nation, citizen, revolution itself.

How it was read

Tocqueville's The Old Regime and the Revolution (1856) noticed something counter-intuitive: revolution broke out in the parts of France where conditions had been improving, not where they were worst. “The most dangerous moment for a bad government is usually when it begins to reform.” This became the idea of rising expectations and relative deprivation. A century later, Theda Skocpol's States and Social Revolutions (1979) compared France, Russia and China and argued that revolutions are not made by revolutionaries but happen when states break down under international pressure and peasants are able to revolt.

Why it matters now Relative deprivation — the gap between what people expect and what they get — is often more explosive than absolute hardship. Technologies that rapidly raise and then frustrate expectations are politically volatile.

The Great Depression

Worldwide, 1929–late 1930s

After the Wall Street crash of October 1929, world trade collapsed and unemployment in the United States and Germany reached roughly a quarter or more of the workforce. In Germany the crisis helped bring the Nazis to power; in the United States it produced the New Deal; in Sweden, the foundations of the welfare state.

How it was read

The Marienthal study (see the studies) showed that unemployment produced apathy rather than revolution. Decades later, Glen Elder's Children of the Great Depression (1974), following Californians born in the 1920s, founded life course research: the same historical event affects people differently depending on the age at which it hits them.

Why it matters now If AI does displace work at scale, Marienthal and Elder tell us to look for withdrawal and lost time-structure as much as anger, and to expect very different effects on workers who are twenty-five, forty-five and sixty.

The Holocaust

Nazi-occupied Europe, 1941–45

The systematic murder of six million Jews, along with Roma, disabled people and others, by the Nazi regime and its collaborators.

How it was read

For decades sociology treated the Holocaust as an aberration, a breakdown of civilisation. Zygmunt Bauman's Modernity and the Holocaust (1989) argued the opposite: the genocide was made possible by modern institutions working normally. Bureaucracy divided the task into small technical steps; each official dealt only with paper; moral responsibility was dissolved in the division of labour; and the victims were turned into categories and numbers. Hannah Arendt's report on the Eichmann trial (1963) had described the “banality of evil” of a functionary who seemed to think only about doing his job. Milgram's experiments were designed in response.

Why it matters now Bauman's argument about moral distance — that the more steps between an action and its consequences, the easier harm becomes — is one of the most important for automated systems. When a decision is split between a data supplier, a model developer, a vendor and an operator, who is responsible?

Decolonisation

Asia, Africa, the Caribbean and the Pacific, c. 1945–75

In three decades, most of the world's colonies became independent states: the Philippines in 1946, India and Pakistan in 1947, Indonesia in 1949, Ghana in 1957, most of French and British Africa around 1960, Algeria after a brutal war in 1962. The number of UN member states more than doubled.

How it was read

Western modernisation theory saw new nations as societies that would follow the path of the West. Thinkers from the colonised world disagreed. Frantz Fanon's The Wretched of the Earth (1961) analysed the psychology and violence of colonialism. Dependency theorists argued that formal independence left economic subordination intact (see Theory and politics). Alatas and others argued that the colonised mind needed decolonising too.

Why it matters now The argument about whether formal independence is enough has been revived for technology: countries may be sovereign on paper but dependent on foreign platforms, cloud infrastructure and AI models for basic functions of government and economy.

The US civil rights movement

United States, c. 1954–68

From the Montgomery bus boycott (1955–56) through sit-ins, Freedom Rides, the March on Washington (1963) and Selma, Black Americans and their allies dismantled legal segregation in the South, winning the Civil Rights Act of 1964 and the Voting Rights Act of 1965.

How it was read

Older theories treated protest as irrational “collective behaviour”, a product of social strain and crowd emotion. The civil rights movement was plainly nothing of the kind, and it transformed the field. Doug McAdam's Political Process and the Development of Black Insurgency (1982) stressed changing political opportunities. Aldon Morris's The Origins of the Civil Rights Movement (1984) showed that the movement was built on existing indigenous organisations — Black churches, colleges and local associations — that provided leaders, communication networks, money and meeting places. McAdam's Freedom Summer (1988) found that the young volunteers who went South in 1964 were distinguished less by their beliefs than by their biographical availability and their ties to others who went.

Why it matters now Movements are built on organisation and relationships, not just grievance or attention. That finding sets up the central puzzle of movements in the social media age (see the Arab Spring, below).

1968

Paris, Prague, Mexico City, Chicago, Tokyo, Berkeley and elsewhere

A year of student revolts, general strikes and uprisings across the world: the May events in France, the Prague Spring and its crushing by Soviet tanks, the massacre of students at Tlatelolco in Mexico City, protests against the Vietnam War. In the Philippines, the late 1960s student radicalism led into the “First Quarter Storm” of early 1970.

How it was read

The new movements did not fit class politics. Alain Touraine and Alberto Melucci called them new social movements: concerned with identity, autonomy, the environment, gender and peace rather than wages. Ronald Inglehart's The Silent Revolution (1977) argued that generations raised in post-war security were shifting from materialist to postmaterialist values, a claim tracked since by the World Values Survey.

Why it matters now Inglehart's later work, with Pippa Norris, argued that the backlash against postmaterialist values helps explain today's populism. Many of the cultural conflicts now fought over AI — about bias, content moderation and whose values models should hold — continue the argument that 1968 started.

The Iranian Revolution

Iran, 1977–79

Months of mass demonstrations and strikes forced the Shah, backed by a powerful army and the United States, to leave the country in January 1979. Ayatollah Khomeini returned and an Islamic Republic was established.

How it was read

It embarrassed the theory of revolutions. Skocpol's structural account, published that same year, said revolutions happen when states collapse under external pressure; Iran's state had not collapsed militarily, and the revolution was made, to a remarkable degree, by an urban movement inspired by religious ideas. Skocpol herself published a notable revision in 1982 acknowledging the role of culture and ideology. Charles Kurzman's The Unthinkable Revolution in Iran (2004) went further: the revolution happened in part because enough Iranians came to believe it would happen — and, he argued, events like this may be inherently unpredictable.

Why it matters now A permanent caution against confident forecasts of social change, including from models trained on the past.

People Power: EDSA, Manila, 1986

Philippines, 22–25 February 1986

After twenty years in power, fourteen of them under martial law and authoritarian rule, Ferdinand Marcos called a snap presidential election for 7 February 1986 against Corazon Aquino, widow of the assassinated opposition leader Benigno Aquino Jr. The official count declared Marcos the winner; the citizens' watchdog NAMFREL's count and many observers said otherwise, and computer technicians at the official canvass walked out in protest at the manipulation of the tally. On 22 February the defence minister Juan Ponce Enrile and the deputy armed forces chief Fidel Ramos broke with Marcos and barricaded themselves in two military camps on Epifanio de los Santos Avenue (EDSA). Cardinal Jaime Sin called on the faithful over Radio Veritas to protect them. Hundreds of thousands, by many estimates over a million, filled the avenue, with nuns and civilians placing themselves in front of tanks. On 25 February, Marcos fled to Hawaii.

How it was read

Sociologists and political scientists have stressed several elements together: a regime weakened by economic crisis and the 1983 assassination; a divided military, in which the reformist faction's defection was decisive; the Catholic Church as a mobilising structure with authority, networks and a radio station; a vigorous civil society, including NAMFREL; and a deliberately nonviolent repertoire that made soldiers unwilling to fire. It became a model, and a name, for the wave of democratic transitions that followed across Asia and Eastern Europe.

The digital echo

In January 2001, the second EDSA uprising that ousted President Joseph Estrada was organised in large part by text messages. Howard Rheingold's Smart Mobs (2002) treated it as one of the first demonstrations of mobile-coordinated collective action, and it became a reference point for later debates about technology and protest.

The darker echo

By the mid-2010s the Philippines, among the world's heaviest users of social media, had become a key site for studying digital disinformation. Jonathan Corpus Ong and Jason Cabañes's study Architects of Networked Disinformation (2018) interviewed the advertising and PR strategists, influencers and fake-account operators behind political campaigns, and found an industry of precarious creative workers rather than a single villain. The journalist Maria Ressa, whose news site Rappler documented these networks, shared the 2021 Nobel Peace Prize. Researchers have also documented sustained online campaigns that recast the martial law era in a positive light.

Why it matters now EDSA shows how institutions, networks and nonviolent discipline turn a crowd into a political force. The decades since show how the same networked publics can be organised, for hire, to rewrite memory. Both lessons apply directly to generative AI, which lowers the cost of producing persuasive content to almost nothing.

The fall of the Berlin Wall

East Germany and Eastern Europe, 1989

Within a few months in 1989, communist regimes that had seemed immovable for forty years collapsed across Eastern Europe, mostly without violence. In East Germany, weekly Monday demonstrations in Leipzig grew from a few thousand people in September to hundreds of thousands in October; on 9 November, after a botched announcement about travel rules, crowds crossed the Wall.

How it was read

Almost no one predicted it. The economist Timur Kuran explained why in “Now Out of Never” (1991) and later Private Truths, Public Lies (1995): under repression, people engage in preference falsification, publicly supporting a regime they privately oppose. Nobody, including the regime and the opposition, knows how deep opposition really runs. Once a few people show their true preferences, others with slightly higher thresholds follow, producing a cascade. This built on Mark Granovetter's threshold models of collective behaviour (1978), in which the same population can produce a riot or nothing depending on small differences in the distribution of thresholds.

Why it matters now Threshold and cascade models are the formal ancestors of how information spreads, or fails to, on social platforms. They also imply that algorithms deciding what each person sees about what others think can tip collective outcomes.

The 2008 financial crisis

United States, then worldwide, 2007–09 and after

The collapse of the US housing bubble, the failure of Lehman Brothers in September 2008, bank bailouts, a global recession and, in Europe, a sovereign debt crisis and years of austerity.

How it was read

Economic sociologists had been studying financial markets as social systems. Donald MacKenzie's An Engine, Not a Camera (2006) argued that financial models are performative: the Black-Scholes option pricing formula did not just describe markets, it changed them until they fitted it, and then helped make them fragile. Neil Fligstein and colleagues showed how the mortgage industry's organisational structure produced the crisis. The aftermath produced Occupy Wall Street in 2011 and its slogan, “We are the 99%”, which put inequality back at the centre of public debate.

Why it matters now Performativity is a key concept for AI: models used for decisions reshape the world they model. When everyone uses similar models, they can fail all at once.

The Arab Spring

Tunisia, Egypt, Libya, Syria, Yemen, Bahrain and elsewhere, 2010–12

The self-immolation of the Tunisian street vendor Mohamed Bouazizi in December 2010 triggered protests that forced President Ben Ali from power on 14 January 2011. Egypt's Hosni Mubarak resigned on 11 February after eighteen days of protest centred on Tahrir Square. Uprisings followed across the region. Tunisia's democratic transition lasted for a decade; elsewhere, outcomes included a military coup in Egypt, civil wars in Libya, Syria and Yemen, and repression in Bahrain.

How it was read

Early commentary called them “Twitter revolutions”. Sociologists were more careful. Zeynep Tufekci's Twitter and Tear Gas (2017) offered the most influential analysis. Digital tools let movements grow very fast and gather huge crowds without the years of organisation the civil rights movement needed. But that speed meant they skipped building the capacities — decision-making structures, trust, the ability to negotiate and to change tactics — that slow organising produces. The result was a “tactical freeze”: movements that could fill a square but could not decide what to do next. Governments, meanwhile, learned to respond not by censoring but by flooding the information space.

Why it matters now Tufekci's distinction between the ability to mobilise and the capacity to act together is one of the most useful ideas for thinking about technology and collective life. AI makes producing messages even cheaper; it does nothing to build trust.

#MeToo

Worldwide, from October 2017

The activist Tarana Burke had used the phrase “Me Too” since 2006 for survivors of sexual violence. After reporting on the film producer Harvey Weinstein in October 2017, the actor Alyssa Milano invited women to reply “me too”; the hashtag was used millions of times within days and spread across dozens of countries in local forms.

How it was read

Sociologists analysed it as a case of networked counterpublics and hashtag activism (Sarah J. Jackson, Moya Bailey and Brooke Foucault Welles, #HashtagActivism, 2020): marginalised groups using platforms to make private experience into public issue, in exactly the sense Mills meant. Research also examined its limits — whose stories were amplified, the role of celebrity — and the backlash.

Why it matters now The same platform features that let #MeToo spread decide which voices are visible. Shifts in moderation and ranking policy can amplify or bury a movement without anyone noticing.

Black Lives Matter, 2020

United States and worldwide, summer 2020

The movement began as a hashtag in 2013 after the acquittal of George Zimmerman for killing Trayvon Martin. The murder of George Floyd by a Minneapolis police officer in May 2020, recorded on a bystander's phone, set off protests in thousands of US towns and many countries. Polling analysed by The New York Times suggested that between 15 and 26 million Americans took part, which would make them among the largest protests in the country's history.

How it was read

Research focused on the role of video as evidence, on the remarkable geographic spread into small and largely white towns, on short-term changes in public opinion on race and policing (which rose sharply and then partly receded), and on the policy results. It also sparked a reckoning about technology: as noted above, several companies halted facial recognition sales to police within weeks.

Why it matters now The rise and partial fall of opinion change after 2020 is a case study in the difference between attention and durable change — a distinction platforms optimised for attention may blur.

COVID-19

Worldwide, 2020–23

The pandemic killed millions of people; the World Health Organization's estimates of excess deaths for 2020–21 alone were around 15 million. It also ran the largest unplanned social experiment in history: lockdowns, school closures, remote work and emergency welfare across nearly every country at once.

How it was read

Sociologists documented a steep social gradient: in many countries, mortality was far higher among poorer people, racial and ethnic minorities, and “essential workers” who could not work from home, while professionals moved online. Researchers studied trust as a predictor of compliance with public health measures and of vaccine uptake; the WHO warned of an “infodemic” of misinformation. The pandemic revealed and deepened gender inequality in unpaid care work, and school closures produced learning losses that fell hardest on poorer children.

Why it matters now The remote work divide was a rehearsal for AI's labour effects: those whose work is done on a screen were transformed first. And it showed that in a crisis, trust in institutions is a matter of life and death — and a resource that information environments can build or destroy.

The climate movement

Worldwide, with a surge from 2018

Greta Thunberg's school strike outside the Swedish parliament in August 2018 grew into Fridays for Future; in September 2019, climate strikes drew millions of people in many countries.

How it was read

The climate movement has drawn heavily on research by the political scientists Erica Chenoweth and Maria Stephan. Their Why Civil Resistance Works (2011) analysed 323 campaigns between 1900 and 2006 and found nonviolent campaigns succeeded about 53% of the time, compared with 26% for violent ones, largely because they attract far broader participation. Chenoweth later noted that no campaign in the data that had drawn active participation from 3.5% of the population at its peak had failed — the “3.5% rule”, which she treats as a descriptive pattern, not a law. Her more recent work finds that the success rate of nonviolent campaigns has fallen markedly since about 2010, possibly because digital mobilisation produces larger but less organised movements, and because regimes have adapted.

Why it matters now The decline in effectiveness fits Tufekci's argument about digital movements. It suggests that the tools that make mobilisation easy may make winning harder.

The generative AI shock and the 2023 strikes

From November 2022

OpenAI released ChatGPT on 30 November 2022. It reached an estimated 100 million users within about two months, then the fastest adoption of any consumer application recorded. Within a year, generative AI had triggered rewritten school assessment rules, lawsuits by authors and artists, emergency policy debates and a new investment boom.

The first major labour conflict over it came quickly. The Writers Guild of America struck from May to September 2023 — 148 days — over pay in the streaming era and the use of AI. The agreement established that AI cannot write or rewrite literary material, that AI-generated text cannot be treated as source material that lowers writers' credit or pay, and that studios cannot require writers to use AI, though writers may choose to. The actors' union SAG-AFTRA struck from July to November 2023 and won requirements for consent and compensation for digital replicas of performers.

How to read it

Sociologically, the release of ChatGPT was an event in Sewell's strong sense: not the technology itself, which built on years of research, but the moment when millions of people began to define the situation differently, and institutions — schools, newsrooms, unions, regulators — had to respond. The strikes are a textbook case of the Luddite question correctly understood: not whether to use a technology, but who controls how it is used and who shares in the gains.

Why it matters now The Hollywood contracts were among the first collectively bargained rules on generative AI anywhere. Whether workers in less organised sectors, and in countries without strong unions, can win anything similar is one of the defining questions of the decade. Continue to The algorithmic society.