Social Facts

Marx asked what divides a society. Émile Durkheim asked what holds one together — and in the course of answering, did something Marx never attempted: he built sociology into a university discipline, with a defined object, a method, a journal, a curriculum and a generation of students.

This lesson is about the concept he put at the foundation of the whole enterprise. Everything else in Topic 4.2 rests on it.

Try to break a very small rule.

Not a law. Nothing that would get you arrested. Just one of the tiny conventions nobody has ever written down.

Stand facing the back wall of a lift, looking at the other passengers, for four floors. Or answer "how are you?" with a full and honest account of how you are. Or, in a queue, stand ten centimetres too close to the person ahead. Or wear formal evening dress to a supermarket on a Tuesday morning.

Now — before doing it — notice what you feel.

There is a resistance. A distinct reluctance. Something that pushes back before you have moved.

Where is that push coming from?

Not from any law. No statute governs which way you face in a lift. Not from a person: the other passengers have made no threats and, most likely, would say nothing. Not from your own preference — you have just decided you would like to try it.

And yet the pressure is entirely real. You can feel it in your body. Most people, asked to do these things, find they cannot without considerable effort, and report afterwards that it was far harder than it sounds.

Durkheim's claim: that pressure is the thing sociology studies. It comes from outside you. It acts on you. It exists whether or not you approve of it. And it does not go away when you understand it — you can know precisely why lift convention exists and still feel the resistance.

He called such things social facts, and he insisted they were as real, and as susceptible to scientific study, as anything a physicist measures.

Durkheim was writing in the 1890s, and his problem was institutional as much as intellectual.

Sociology at that point had a name (3.3.1), some grand systems, and no secure place in any university. To established academics it looked like a pretension: everything it claimed to study, they said, belonged to someone else.

"You study individuals, so you are doing psychology." Society is just people; explain the people and you have explained everything.

"You study the past, so you are doing history."

"You study what people ought to do, so you are doing moral philosophy."

Durkheim's whole strategy was to answer with a distinct object of study — something real, observable, and irreducible to any other science's subject matter. Without that, sociology would remain a branch of something else and would never get a chair, a department or a degree.

The Rules of Sociological Method (1895) is that answer, and its first chapter is titled "What is a Social Fact?"

The strategy worked. Durkheim got France's first chair in the subject, founded the journal L'Année Sociologique in 1898, trained a school, and put sociology on the national curriculum for teacher training — which meant every French schoolteacher for a generation learned it. Marx changed the world; Durkheim changed the syllabus, and in the long run that is why sociology exists as a department.

The idea, in plain words

Durkheim's definition, from The Rules:

"A social fact is every way of acting, fixed or not, capable of exercising on the individual an external constraint; or again, every way of acting which is general throughout a given society, while at the same time existing in its own right independent of its individual manifestations."

Unpacked, it has three properties.

1. Externality. Social facts exist outside any particular individual. The French language, the rules of contract, the institution of marriage, the currency, the practice of queueing — none of these was created by you, none depends on you, and all of them were there before you arrived and will remain when you go.

2. Constraint. They exercise pressure on individuals. Sometimes visibly, through law and punishment; far more often invisibly, through ridicule, awkwardness, exclusion, or simply the impossibility of doing otherwise — try inventing your own private language and see how far communication gets.

3. Generality with independence. They are widespread in a society and they exist in their own right, over and above their individual instances. The suicide rate of a country is not any individual's suicide; the marriage rate is not any couple's marriage. These are properties of the collective, not summaries of individuals.

This is Durkheim's most famous methodological instruction, and the most misread. In French: considérer les faits sociaux comme des choses.

He does not say social facts are things — that they are physical objects. He says treat them as things, which is an instruction about method, not a claim about ontology.

What the instruction means, in practice:

Study them from outside, as an object of investigation rather than through introspection. You cannot discover the rules of your society by thinking hard about your own feelings, any more than you could discover the chemistry of water by contemplating thirst.

Discard preconceptions. Durkheim's rule: begin by defining a phenomenon by its external, observable characteristics, not by what everyone assumes it to be. His definition of crime, for instance, is not "acts that are harmful" but "acts that provoke punishment" — an external criterion you can apply without smuggling in your own moral judgement.

Look for causes in other social facts, not in individual motives. And here is his hardest rule, from The Rules: "The determining cause of a social fact must be sought among antecedent social facts, and not among the states of the individual consciousness."

Expect them to resist you. This is the mark of the real: a thing pushes back. Try to change your society's language, calendar or currency by an act of will and you will discover their thinghood immediately.

Why society is not reducible to individuals

The argument beneath the whole programme, and Durkheim's answer to "you are just doing psychology." He calls it emergence, or in his own phrase, that society is sui generis — of its own kind.

This is one of the most consequential arguments in the discipline's history. It is the warrant for sociology existing at all (1.4.1), it is inherited from Comte's hierarchy of the sciences (3.3.1), and it remains the central battlefield against methodological individualism (5.3, 7.2). Whether the argument fully works is a live question; that it founded a discipline is not.

The clearest single instance, and Durkheim uses it himself.

External? Entirely. You were born into a language already fully formed, with a vocabulary, a grammar and a set of distinctions you had no part in choosing.

Constraining? Absolutely. You may not put words in any order you like. If you say "milk the bought I yesterday," you will not be arrested — but you will not be understood, which is a more total constraint than any law.

General and independent? The language exists over and above every sentence anyone has ever spoken in it. Every speaker could die tomorrow and the language, in books and recordings, would exist as an object.

And now the deep part. The language does not merely constrain what you say. It shapes what you can readily think and notice. It gives you the categories — the distinctions between colours, relatives, times, emotions — that come pre-installed. A language with distinct words for your mother's brother and your father's brother makes a distinction that English-speakers must labour to keep in view.

So: something entirely outside you, imposed on you without consent, shapes the interior of your thought. That is the whole Durkheimian programme in one example. And it explains why he thought "society is inside you" (1.1.3) was not a metaphor.

Durkheim distinguishes several, and the distinctions matter.

Morphological (or structural) facts — the "anatomical" features of a society: population size and density, geographical distribution, the arrangement of settlements, the channels of communication. These are the most thing-like, and Durkheim gives them causal priority. Dynamic density — the frequency and intensity of interaction among a population — is the key variable in The Division of Labour (4.2.2).

Institutional facts — established, crystallised ways of acting: law, religion, marriage, contract, property, education systems.

Currents of opinion (courants sociaux) — less fixed, more fluid: enthusiasms, panics, waves of feeling. Durkheim's examples include the surge of collective feeling in a crowd, and the rates of suicide, marriage and migration that fluctuate with the moral temper of a society. These are the least thing-like, and the most interesting: they show that a social fact need not be solid to be real.

And "normal" versus "pathological" facts — a distinction that gets him into trouble. Durkheim held that a social fact is normal when it is general in a society of that type at that stage of development, and pathological when it deviates. His most startling application: crime is normal. It occurs in every society; a society without crime is impossible, since if serious offences vanished, minor deviations would attract the same outrage; and crime is even useful, since it provokes the collective reaction that reaffirms shared standards, and since today's crime is sometimes tomorrow's morality — Socrates, he notes, was a criminal.

The trouble is that "normal = general" easily becomes "general = healthy = right," which smuggles a value judgement into what was supposed to be a value-free method. Critics have made this objection since 1895 and it has never been satisfactorily answered.

Go deeper

The three great objections

One: reification. Does Durkheim treat "society" as a real entity with its own existence, above and beyond the people in it — a group mind, a thing that acts and wants? Critics from Gabriel Tarde (his contemporary and most persistent opponent) onward have said yes. Durkheim's defence is that a social fact exists in individual consciousnesses collectively, not in some free-floating spirit; the conscience collective (4.2.5) is a distribution across minds, not a mind of its own. The defence is not always convincing, and some of his own formulations invite the charge.

Two: where is the individual? If social facts determine behaviour, what becomes of choice, resistance, creativity, interpretation? This is the structure/agency problem (1.3.1) in its sharpest form. Durkheim's actual position is more nuanced than his slogans: he insists constraint is not the whole story, that people internalise social facts and experience them as their own obligations rather than as external chains, and that in modern societies individual autonomy is itself a social fact — a collectively produced value, the "cult of the individual" (4.2.5). But the framework tilts heavily towards structure, and Weber's insistence on meaning and interpretation (4.3.1) is the direct counterweight.

Three: too much constraint, not enough conflict. Whose social facts? A rule that feels natural to those it benefits may be experienced as imposition by those it does not. Durkheim tends to write as if a society's moral rules were shared by everyone in it — where Marx would ask whose interests they serve, and Weber would ask which groups fought over them. The conflict tradition's standing criticism of the whole Durkheimian line (5.2) starts here.

The methodological rules

Beyond the definition, The Rules lays out procedures that are still, in modified form, standard.

Define phenomena by external, observable characteristics before investigating them, so that everyday prejudice does not do the work in advance.

Eradicate preconceptions (prénotions) — the ready-made ideas we all carry about family, crime, religion. This is 1.4.6 on common sense, in its founding statement.

Distinguish cause from function. "When we undertake to explain a social phenomenon, we must seek separately the efficient cause which produces it and the function it fulfils." A thing's usefulness does not explain its origin. This is a warning against functionalist over-reach (5.1) issued by the person usually blamed for it — which is worth noticing.

Use the comparative method, and specifically concomitant variation — comparing cases in which a supposed cause is present and absent. Since sociology cannot experiment, comparison is its substitute. Suicide (4.2.4) is the demonstration.

One cause, one effect. If a phenomenon has different causes in different cases, Durkheim argues, it is not one phenomenon but several. This rule is now generally rejected — social outcomes are routinely multiply caused — but it explains why Suicide is structured as it is.

The lasting form of the idea

Where the concept survives today, usually without the name attached.

Institutions as objects. The entire field of new institutionalism — the study of how organisational forms, once established, constrain and shape the people inside them — is Durkheimian in structure. Berger and Luckmann's The Social Construction of Reality (2.4.1) explicitly builds on him: institutions are humanly produced and then experienced as objective reality, a process they call objectivation.

Rates as evidence. Every use of official statistics to reveal something about a society rather than about its individuals rests on Durkheim's argument that a rate is a property of a collective. Every study showing that a country's homicide rate, fertility rate or life expectancy tracks its institutional arrangements is doing what he said sociology should do.

Emergence beyond sociology. The argument that wholes have properties their parts lack is now standard in complexity science, biology and philosophy of mind. Durkheim's version was early, and cruder than modern statements of it, but recognisably the same argument.

And the sharpest modern test: language and technology. A social media platform's affordances are external to any user, constrain what can be done, exist independently of any individual, and shape how millions think and feel. They are also owned by someone, which Durkheim's framework has no easy way to register — and which is exactly where Marx's questions have to be brought back in.

"Social facts are physical things." Durkheim says treat them as things — a rule of method, not a claim that they are made of matter.

"Durkheim denied individual agency." He argued that constraint is largely internalised and experienced as obligation, and that individual autonomy is itself a social product. The framework nonetheless leans hard towards structure.

"Social facts are just what most people do." Generality alone is not enough. Everyone sleeps and eats; those are biological. A social fact must also be external and constraining, and must exist independently of its individual instances.

"'Treat as things' means ignore meaning." Durkheim's own greatest work is about religious meaning (4.2.6). The rule is about avoiding introspection and preconception, not about ignoring what things mean to people.

"Crime is normal' means crime is good." He means it is universal, structurally inevitable, and serves the function of provoking moral reaffirmation. He is not endorsing it — though the normal/pathological distinction is his weakest move and does slide towards endorsing whatever is common.

"Durkheim invented functionalism." He explicitly warned that function does not explain cause. Later functionalists (5.1) frequently ignored that warning and cited him anyway.

Why it matters

It is sociology's declaration of independence. Without a distinct object of study, sociology is a branch of psychology, history or philosophy. The social fact is that object, and the argument secured the discipline a place in universities that it has not lost.

It gives the "invisible pressure" a name. Lesson 1.1.2 said social structure is hard to see. Durkheim explains why: the strongest social facts are precisely the ones that feel most natural, because we have internalised them so thoroughly that their constraint registers only when we try to break them. The lift experiment is the demonstration.

It legitimates the study of rates. Suicide rates, crime rates, birth rates — treated not as tallies of individual decisions but as properties of a collective, with social causes. This is the foundation of quantitative sociology, and 4.2.4 is the proof of concept.

And it sets up the opposition that organises Part 5. Durkheim's insistence on the primacy of structure and the reality of the collective; Weber's insistence that action must be understood through the meanings actors give it (4.3.1). Almost every theoretical dispute in the twentieth century can be located somewhere between those two positions.

A social fact is a way of acting, thinking or feeling that is external to the individual, exercises constraint, and is general in a society while existing independently of its individual manifestations.

Durkheim's instruction — "treat social facts as things" — is methodological, not ontological: study them from outside, discard preconceptions, define by external criteria, and expect them to resist you.

The key rule: "The determining cause of a social fact must be sought among antecedent social facts, and not among the states of the individual consciousness."

The underlying argument is emergence — society is sui generis, as bronze is hard though copper and tin are soft. This is why explaining individuals will never explain society, and it is sociology's warrant for existing.

Kinds: morphological (population, density, dynamic density), institutional (law, marriage, property), and currents of opinion — plus the troublesome normal/pathological distinction, under which crime is normal and even functional.

Objections: reification (does "society" become an entity?); the disappearance of agency; and the absence of conflict — whose social facts?

Other rules: eradicate prénotions; separate cause from function (a warning against functionalism, from the man blamed for it); use the comparative method and concomitant variation.

The strategy was institutional as well as intellectual: France's first chair, L'Année Sociologique (1898), a school of students, and sociology in teacher training. Marx changed the world; Durkheim changed the syllabus.

Social fact (fait social) — external, constraining, general and independent ways of acting, thinking and feeling.

Externality — existence outside and prior to any particular individual.

Constraint — the pressure a social fact exerts, whether by law, sanction, ridicule or sheer impossibility of doing otherwise.

Sui generis — "of its own kind"; Durkheim's claim that social reality is a distinct level, not reducible to individual psychology.

Emergence — the appearance of properties at the level of a whole that no part possesses.

"Treat social facts as things" — the methodological rule of studying social phenomena from outside, free of preconception.

Prénotions — ready-made everyday ideas that must be cleared away before investigation.

Morphological facts — the structural features of a society: size, density, distribution.

Dynamic density — the frequency and intensity of interaction in a population; Durkheim's key causal variable.

Currents of opinion (courants sociaux) — fluid collective states such as enthusiasms, panics and rates.

Normal and pathological — Durkheim's distinction between what is general for a social type and what deviates; his most criticised move.

Concomitant variation — the comparative method of examining cases where a supposed cause is present and absent.

One — do the lift experiment. Or a milder one: stand slightly too close in a queue, or say "no, actually" when asked how you are. Note the resistance before you act, and the reactions after. You are not being naughty; you are measuring a social fact by its constraint. (Do nothing that would frighten or harm anyone — Garfinkel's breaching experiments in 2.4.3 raised exactly this ethical question.)

Two — list ten facts you did not choose. Language, calendar, currency, measurement units, the side of the road, the working week, the alphabet's order, the shape of a plug, the age of majority, the ritual of a handshake. For each, ask: could you change it alone? The answer is the definition.

Three — apply the external definition. Define "crime," "family" or "success" by externally observable characteristics, without reference to what you think it should be. Notice how hard this is. That difficulty is what prénotions means.

Four — find a rate. Look up any national rate — road deaths, marriage, fertility, homicide — for your country over thirty years. It will be remarkably stable year to year, then move in trends. Individual decisions are wildly variable; the rate is not. That stability is Durkheim's evidence.

Five — test the reification objection. Write the sentence "society requires X of us" and then rewrite it naming actual people, institutions and mechanisms. Which version is more informative? Do this honestly — sometimes the first is a useful shorthand and sometimes it hides everything that matters.