How a science of society came to be

§ 1Before the word existed

People have always theorised about society. What changed in the nineteenth century was the ambition to study it systematically, with evidence, as something that has its own laws rather than as a reflection of God's will or a king's. But the ambition had ancestors, and one of them deserves more than a footnote.

In 1377, in a castle in what is now Algeria, the historian Ibn Khaldun finished the Muqaddimah, the introduction to a universal history. In it he proposed that dynasties rise and fall in a regular cycle driven by asabiyya, usually translated as group feeling or social cohesion. Desert tribes, bound by kinship and hardship, have strong asabiyya and conquer settled cities. Within three or four generations the conquerors grow used to luxury, depend on paid soldiers and bureaucrats, lose their cohesion, and are conquered in turn. It is a theory of social change based on solidarity, economic life and the state, and it anticipates Durkheim's questions by five centuries.Arnold Toynbee called the Muqaddimah “the greatest work of its kind that has ever yet been created by any mind in any time or place”. That is overblown, but the neglect of Ibn Khaldun in European-centred histories of sociology was a real distortion, one that Syed Farid Alatas and others have spent careers correcting.

Closer to the founders, the European Enlightenment produced a cluster of proto-sociologists. Montesquieu's The Spirit of the Laws (1748) argued that a society's laws fit its climate, economy and customs. The Scottish thinkers Adam Ferguson and Adam Smith described how a commercial society organises itself without anyone designing it — the unintended consequences of individual action, a theme sociology would never abandon. Alexis de Tocqueville's Democracy in America (1835 and 1840) is still assigned in sociology courses for its account of voluntary associations, equality and the “tyranny of the majority”.

§ 2The founders' problem

Sociology proper came out of what the historian Eric Hobsbawm called the dual revolution: the political revolution that began in France in 1789 and the industrial revolution that began in Britain a little earlier. Between them they destroyed the old order. Peasants became wage workers, villages emptied into cities, the church lost its hold on public life, and inherited rank gave way to contracts and markets.

Every classical sociologist was, in one way or another, asking the same question: what holds a society together once tradition stops doing it? Marx answered that nothing really does — capitalism is held together by force and illusion until it breaks. Durkheim answered that a new, more complex kind of solidarity grows out of the division of labour, if it is allowed to. Weber answered that what replaces tradition is rational calculation, and that it may become a cage.

Hold onto that question. It is the same one being asked about AI: what happens to trust, work, authority and meaning when a new technology dissolves the arrangements people relied on?

§ 3Auguste Comte and Harriet Martineau

Auguste Comte (1798–1857) gave the field its name in 1838, in the fourth volume of his Course of Positive Philosophy.The word had appeared earlier in an unpublished manuscript by the Abbé Sieyès around 1780, but Comte gave it currency. He had originally called the field “social physics” and switched when he found the Belgian statistician Adolphe Quetelet using that phrase. Comte believed human thought passed through three stages — theological, metaphysical, and finally “positive” or scientific — and that a science of society would sit at the top of the hierarchy of sciences, able to guide society towards order and progress. In his later years he founded a secular Religion of Humanity with its own calendar of saints. The motto on Brazil's flag, Ordem e Progresso, is his.

Comte's specific claims have not aged well. His lasting contribution is the idea that society can be studied with the same seriousness as nature, and the separation of “social statics” (how society holds together) from “social dynamics” (how it changes).

Harriet Martineau (1802–1876) did more to make that idea practical. Her How to Observe Morals and Manners (1838) is arguably the first book on sociological method: it tells the traveller to study a society through its institutions, records and things rather than through conversation with elites, and to suspend judgement. Her Society in America (1837) tested the country against its own professed ideals and found it wanting on slavery and the position of women. In 1853 she translated and condensed Comte into English, and Comte liked her version so much that he had it translated back into French.

§ 4Karl Marx

Karl Marx (1818–1883) did not call himself a sociologist, but no single thinker has shaped the field more. His starting point was material: to understand a society, look first at how it produces what it needs to live, and at who owns the means of producing it. The relations of production — lord and serf, owner and worker — shape law, politics, religion and ideas, which he called the superstructure.

Under capitalism, workers sell their labour power for a wage worth less than the value they produce; the difference, surplus value, is profit. Capitalism is therefore built on exploitation, even when every contract is freely signed. It also produces alienation: workers are estranged from what they make, from the work itself, from each other and from their own creative nature. Its drive to accumulate makes it endlessly dynamic and periodically self-destructive. “All that is solid melts into air,” as the Communist Manifesto (1848, with Friedrich Engels) put it.

Marx's predictions of proletarian revolution in the industrial West did not come true, and the states built in his name became some of the most repressive of the twentieth century. But his questions about class, ownership, technology and power have never left sociology. In a notebook passage from 1857–58 later known as the “Fragment on Machines”, he imagined a future in which the “general intellect” — society's accumulated knowledge — is embodied in machines and becomes the main force of production, so that labour time stops being the measure of wealth.The passage is in the Grundrisse, notebooks not published until 1939–41. It became a touchstone for Italian autonomist Marxists in the 1970s and again for writers on “cognitive capitalism” and AI. Whether it is prophecy or just a striking paragraph is very much a live argument. That passage is now quoted constantly in debates about large language models, which are, quite literally, machines trained on the general intellect.

§ 5Émile Durkheim

Émile Durkheim (1858–1917) did more than anyone to make sociology an academic discipline. He taught the first social science course in a French university at Bordeaux in 1887, founded the journal L'Année sociologique in 1898, and set out the rules of the game in The Rules of Sociological Method (1895): treat social facts as things. A social fact is a way of acting, thinking or feeling that is external to the individual and exerts constraint — language, law, money, fashion, the suicide rate. It can only be explained by other social facts, not by individual psychology.

In The Division of Labour in Society (1893) he distinguished two kinds of solidarity. In simple societies people are alike, and they are held together by mechanical solidarity, a shared conscience enforced by punishment. In complex societies people are different and specialised, and they are held together by organic solidarity, mutual dependence, like the organs of a body. The danger in the transition is anomie: normlessness, a state in which the old rules have lapsed and new ones haven't formed, leaving people without guidance on what to want or how much is enough.

Then he tested the idea on the hardest possible case. Suicide (1897) used official statistics from across Europe to show that suicide rates were stable within societies and varied systematically between groups — by religion, marital status, and economic conditions. If the most individual act imaginable has social causes, then everything does. (The study and its problems are covered in detail in the studies section.)

His last major book, The Elementary Forms of Religious Life (1912), argued that religion is society worshipping itself: in ritual, a group experiences its own collective power as something sacred. The concept of collective effervescence, the charged feeling of a crowd acting together, came from that book. It is a good description of a viral moment online.

§ 6Max Weber

Max Weber (1864–1920) disagreed with both Marx and Durkheim on method. For him, sociology had to understand the meaning actions have for the people performing them — Verstehen, interpretive understanding — as well as explain their causes. Ideas were not just a superstructure: they could change history.

His most famous demonstration was The Protestant Ethic and the Spirit of Capitalism (1904–05). Calvinist doctrine taught that God had already chosen who would be saved, and no one could know for sure. Believers relieved the anxiety by working relentlessly and living frugally, taking worldly success as a sign of grace. The result was a disciplined, reinvesting, methodical way of life that happened to be perfect for capitalism. Once capitalism was established, it no longer needed the religion. The Puritan, Weber wrote, wanted to work in a calling; we are forced to. The modern economic order had become a stahlhartes Gehäuse — a shell as hard as steel, translated by Talcott Parsons as the “iron cage”.

Behind this is Weber's master theme: rationalisation, the spread of calculation, rules and efficiency into every sphere of life, and the resulting “disenchantment of the world”. Its most complete form is bureaucracy: hierarchical offices, written rules, impersonal procedures, appointment by expertise. Weber thought bureaucracy was technically superior to every other form of administration and also deeply dangerous, because it could escape political control and turn people into “specialists without spirit”.

Weber also gave political sociology its basic vocabulary: the state as the body that successfully claims a monopoly on the legitimate use of physical force within a territory, and three types of legitimate authority — traditional, charismatic and legal-rational. And he insisted that scholars separate facts from values while admitting that values decide which questions get asked.

§ 7Georg Simmel

Georg Simmel (1858–1918) never got a proper professorship in Berlin, partly because of antisemitism and partly because his essays were too brilliant and too unsystematic for the German academy. He studied the forms of social interaction rather than its contents: what happens to a relationship when a third person enters it (the dyad becomes a triad, and coalitions become possible); the social type of the stranger, who is in a group but not of it; secrecy; fashion; the meal.

The Philosophy of Money (1900) argued that money turns qualities into quantities and makes every relationship potentially impersonal and calculable. “The Metropolis and Mental Life” (1903) described how city dwellers, bombarded with stimuli, develop a blasé attitude: an intellectual, reserved, slightly indifferent way of coping with too much. It reads now like a description of the infinite feed.

§ 8Du Bois, Addams and the missing founders

For most of the twentieth century, the canon was three dead European men. That story left out people who were doing rigorous empirical sociology at the same time, often more rigorously.

W. E. B. Du Bois (1868–1963) was the first African American to earn a Harvard PhD. His The Philadelphia Negro (1899) was a door-to-door survey of the city's Seventh Ward that combined statistics, maps and interviews to show that the conditions of Black Philadelphians were the product of discrimination and history, not innate character. At Atlanta University from 1897 he ran an annual series of empirical studies of Black life, now recognised as one of the first American schools of scientific sociology.The sociologist Aldon Morris made the case in The Scholar Denied (2015) that Du Bois and the Atlanta school, not Chicago, founded American scientific sociology, and that they were written out of the history because of race. In The Souls of Black Folk (1903) he described double consciousness: “this sense of always looking at one's self through the eyes of others”, the experience of being both American and Black in a country that treated those as opposites. He declared that “the problem of the Twentieth Century is the problem of the color-line”.

Jane Addams (1860–1935) co-founded Hull-House in Chicago in 1889, a settlement house where educated residents lived among immigrant workers. Its residents produced Hull-House Maps and Papers (1895), detailed maps of nationality and wages block by block, years before the university sociologists down the road made mapping their signature. Addams won the Nobel Peace Prize in 1931. She was gradually pushed out of academic sociology as it professionalised and redefined her kind of work as “social work”, a field coded as female and applied.

Other names belong here too: Ida B. Wells's statistical investigations of lynching in the 1890s; Anna Julia Cooper's A Voice from the South (1892); and, much further back, the Filipino reformer José Rizal's essay “On the Indolence of the Filipino” (1890), which argued that the supposed laziness of colonised people was a product of colonial conditions, not of race — an argument Syed Hussein Alatas would build into a full sociology eighty-seven years later.

§ 9Chicago and the city as laboratory

The University of Chicago opened the first American department of sociology in 1892, and the American Journal of Sociology followed in 1895. In the 1910s and 1920s, under Robert Park and Ernest Burgess, the department turned its own city into a laboratory. Chicago had grown from a few thousand people in 1840 to well over two million by 1920, much of it through immigration, and the sociologists set out to map it.

Burgess's concentric zone model (1925) described the city as a series of rings: business district, a “zone in transition” of factories and slums, working-class homes, then middle-class residences and commuter suburbs. The approach, called human ecology, treated neighbourhoods like ecological niches where groups competed, invaded and succeeded each other. The students produced a famous run of ethnographies: The Gang, The Hobo, The Ghetto, The Taxi-Dance Hall, The Gold Coast and the Slum.

W. I. Thomas and Florian Znaniecki's The Polish Peasant in Europe and America (1918–20) used letters and life histories to study migration. And W. I. Thomas, with Dorothy Swaine Thomas, gave sociology one of its most quoted sentences in 1928: “If men define situations as real, they are real in their consequences.” This is the Thomas theorem, and it is the foundation of everything later written about self-fulfilling prophecies — including predictive models that bring about what they predict.

From the same Chicago tradition came symbolic interactionism, named by Herbert Blumer in 1937 and built on George Herbert Mead's lectures: the self is formed through interaction, by learning to see ourselves as others see us, and social life is a continuous process of interpreting and negotiating meanings.

§ 10Mid-century: system, critique, imagination

Parsons, Merton and functionalism

After the Second World War, American sociology became the world's dominant sociology, and its dominant theorist was Talcott Parsons at Harvard. In The Structure of Social Action (1937) he synthesised Weber, Durkheim and others; in The Social System (1951) he built an enormous framework in which every society must solve four functional problems (adaptation, goal attainment, integration and latency, or AGIL) and every institution contributes to the stability of the whole. Structural functionalism was an attempt at a general theory of everything social. Critics said it couldn't explain conflict or change, and that it read American suburban life in the 1950s as the natural endpoint of history.

Robert K. Merton, Parsons's student, preferred “theories of the middle range” that could actually be tested. He distinguished manifest from latent functions (the intended and unintended consequences of an institution), developed strain theory, which explained deviance as a response to a gap between culturally approved goals and the legitimate means of reaching them, and named the self-fulfilling prophecy (1948): a false definition of a situation that evokes behaviour which makes it come true. His example was a bank run.

The Frankfurt School

The Institute for Social Research was founded in Frankfurt in 1923 and fled the Nazis to New York in the 1930s. Max Horkheimer, Theodor Adorno, Herbert Marcuse and their colleagues tried to explain why the revolution Marx predicted had not happened and why workers had instead embraced fascism or consumerism. Their answer, critical theory, combined Marx with Freud and Weber. In Dialectic of Enlightenment (1944/1947), Horkheimer and Adorno argued that Enlightenment reason had turned into instrumental reason — reason as a tool for domination — and that the “culture industry” produced standardised entertainment that kept people passive. Marcuse's One-Dimensional Man (1964) became a text of the New Left.

C. Wright Mills

C. Wright Mills, at Columbia, attacked both Parsons (“grand theory” lost in abstraction) and survey research (“abstracted empiricism” with no big questions). The Power Elite (1956) argued that America was run by an interlocking group of corporate, military and political leaders. The Sociological Imagination (1959) gave the discipline its best-known self-description: the ability to grasp the relationship between biography and history, between “the personal troubles of milieu” and “the public issues of social structure”.

When, in a city of 100,000, only one man is unemployed, that is his personal trouble… But when in a nation of 50 million employees, 15 million men are unemployed, that is an issue, and we may not hope to find its solution within the range of opportunities open to any one individual.C. Wright Mills, The Sociological Imagination (1959)

§ 11The 1960s break

The civil rights movement, Vietnam, decolonisation and the student revolts of 1968 broke functionalism's hold. If society was a smoothly integrated system, why was it on fire? Several traditions rose at once.

• Conflict theory — Ralf Dahrendorf, Lewis Coser, Randall Collins — put power, interests and struggle at the centre.
• Micro-sociology flourished. Erving Goffman's The Presentation of Self in Everyday Life (1956; US edition 1959) analysed social life as performance, with front stages and back stages. Harold Garfinkel's Studies in Ethnomethodology (1967) showed, often by having students deliberately break everyday rules, how much invisible work goes into producing ordinary order.
• Peter Berger and Thomas Luckmann's The Social Construction of Reality (1966) argued that everything we take as objective reality is produced by people, then hardens into institutions that appear natural, then is internalised by the next generation.
• Labelling theory (Howard Becker's Outsiders, 1963) argued that deviance is not a quality of an act but the result of others successfully applying a label to it.
• Feminist sociology asked why half of humanity had been treated as a footnote. Dorothy Smith argued in 1974 that sociology was written from the standpoint of men who managed institutions, and proposed an “institutional ethnography” starting from women's everyday experience. Ann Oakley's The Sociology of Housework (1974) treated domestic work as work.
• Black sociology challenged the field's assumptions directly, notably in Joyce Ladner's edited collection The Death of White Sociology (1973).

§ 12The grand syntheses

From the mid-1970s, a generation of European theorists tried to overcome the split between structure (society shapes people) and agency (people shape society). Their work remains the toolkit most often used to analyse digital technology.

Michel Foucault's Discipline and Punish (1975) traced the shift from public torture to the prison and argued that modern power works less by force than by surveillance, examination and the production of “normal” individuals. His image of the panopticon, Jeremy Bentham's prison design in which inmates can always be seen but never know when, became the default metaphor for digital surveillance — perhaps too much so.

Pierre Bourdieu's Distinction (1979) showed how taste in food, music and art marks and reproduces class. He gave us habitus (embodied dispositions acquired through upbringing), field (a structured arena of competition) and several forms of capital: economic, cultural, social, and symbolic.

Jürgen Habermas, heir to the Frankfurt School, traced the rise and decline of the public sphere in 1962 and later (1981) built a theory of communicative action, in which legitimate norms emerge from free, reasoned discussion among equals. Every argument about whether social media or AI has damaged democracy is, at some level, an argument with Habermas.

Anthony Giddens's theory of structuration (1984) held that structures are both the medium and the outcome of the practices they organise. Niklas Luhmann's systems theory treated society as a network of self-reproducing communication systems — law, economy, science, mass media — each operating on its own code; it is unusually well suited to thinking about non-human communicators. And Bruno Latour, Michel Callon and John Law developed actor-network theory, which refused to separate the social from the technical and treated objects, from scallops to door closers, as actors in networks.

§ 13Globalisation and the network society

By the 1990s, the nation-state was looking like the wrong unit of analysis. Immanuel Wallerstein's world-systems theory (from 1974) had already described a single capitalist world-economy divided into core, semi-periphery and periphery. Saskia Sassen's The Global City (1991) showed how New York, London and Tokyo had become command posts for a global economy, with a polarised workforce of financiers and the cleaners, couriers and cooks who serve them.

Ulrich Beck's Risk Society (published in German in 1986, the year of Chernobyl) argued that modern societies are increasingly organised around the distribution of risks they have manufactured themselves, risks that are global, invisible and impossible to insure. Zygmunt Bauman described Liquid Modernity (2000): a world where nothing holds its shape for long.

Most directly relevant here, Manuel Castells's trilogy The Information Age (from 1996) described the network society: a social structure made of information networks, in which power lies in programming networks and switching between them, and in which the “space of flows” of capital and information comes apart from the “space of places” where people actually live.

§ 14Sociology from the South

A persistent criticism of the discipline is that it took the experience of a handful of North Atlantic countries, called it “modernity”, and treated everywhere else as a case of not having caught up yet. Raewyn Connell's Southern Theory (2007) made the argument systematically, and recovered social thought from Iran, Africa, Latin America, India and Australia. But the critique had long been made from the South itself.

The Malaysian sociologist Syed Hussein Alatas diagnosed the “captive mind” in the early 1970s: an intellectual dominated by Western thought, imitating its categories uncritically. His The Myth of the Lazy Native (1977) showed how colonial administrators in the Philippines, Malaya and Java constructed an image of lazy natives to justify forced labour and colonial rule — a direct influence on Edward Said's Orientalism, and an explicit continuation of Rizal's argument. In China, Fei Xiaotong's From the Soil (1947) contrasted the Western “organisational mode of association” with China's “differential mode of association”, in which each person sits at the centre of rings of relationships spreading outward like ripples from a stone. In Latin America, dependency theorists (see Theory and politics) and later Aníbal Quijano's concept of the coloniality of power (2000) argued that race and the global division of labour were forged together in the conquest of the Americas.

In the Philippines, the Philippine Sociological Society was founded in 1952 and its journal, the Philippine Sociological Review, began the following year. Debates over indigenisation — whether Filipino social science should build its concepts from local language and experience rather than importing them — ran through the 1970s and after, most famously in the neighbouring field of psychology with Virgilio Enriquez's Sikolohiyang Pilipino.

These traditions matter for AI in a specific way. The systems now deployed worldwide are trained mostly on English-language text produced in the global North, and built by a handful of companies in two countries. The “captive mind” problem has a new and literal form. So does the question of whose labour builds the system: much of the data labelling and content moderation that makes AI usable is done in Kenya, the Philippines, India and Venezuela. See The algorithmic society.

§ 15The digital turn

Sociology noticed computers early. Daniel Bell's The Coming of Post-Industrial Society (1973) predicted that theoretical knowledge would become the central resource of the economy and that professional and technical workers would become the leading class. Shoshana Zuboff's In the Age of the Smart Machine (1988) studied paper mills and offices being computerised and saw two possible paths: technology that automates (replaces people) and technology that informates (generates information that could empower them). Thirty years later her The Age of Surveillance Capitalism (2019) concluded that the industry had chosen a third option she hadn't foreseen.

The real methodological shift came with data. In 2009, fifteen researchers led by David Lazer published a short piece in Science titled “Computational Social Science”, arguing that the digital traces of everyday life — phone records, emails, online transactions — would transform the study of human behaviour, but that most of this data was held by companies and governments. Both halves of the prediction came true. Since then “digital sociology” (Deborah Lupton, 2015; Noortje Marres, 2017) has developed as a field studying both digital society and digital methods, and Matthew Salganik's Bit by Bit (2018) became the standard guide to doing social research in the digital age.

The arrival of large language models in the 2020s changed things again: AI is now both an object of sociological study and an instrument of it. That is the subject of section 06.