The first time you walk into an unfamiliar setting — an expensive restaurant, someone else's company meeting room, a family dinner where everyone knows everyone but you — nobody hands you a manual, yet within seconds you have adjusted: how fast you walk, how loud you talk, where your hands go, when you laugh. Afterwards you could not say which rule you were following.
This page takes apart three things: how that unsayable equipment gets loaded into a body; how institutions like schools and companies read it back as "personal quality"; and why, when it comes to taste, what counts as overdone locates a person better than what counts as good.
Pierre Bourdieu started not from theory but from two stretches of fieldwork. In the late 1950s he did anthropological research in the Kabylia highlands of Algeria; back in France in the early 1960s he returned to Béarn, the rural south-west where he grew up, to work out something the locals themselves found puzzling: eldest sons could not find wives.
By local inheritance custom the farm went to the eldest son, which by that same custom should have made him the best available match. But urbanisation accelerated through the 1950s, young women increasingly preferred to marry into town, and the men the rules favoured most were the ones left single in large numbers. What Bourdieu noticed was this: if "the rules" explained conduct, that outcome could not have arisen.
What actually operated was another layer — which match was thinkable, who at a village dance dared ask whom, what a farmer's way of walking and standing meant to a girl raised in town. Not one of these was written down anywhere, and together they decided who married. His 1962 article on celibacy in Béarn was the first place he used the word habitus, explicitly borrowing from the anthropologist Mauss and his essay on techniques of the body: even how you walk is learned.
Habitus is not a list of behaviours but a generator: a system of embodied dispositions laid down by a durable position — posture, accent, intuitions about what is worth wanting, a sense of what feels awkward. Its key property is that it is generative: in a situation you have never met, it still improvises a response for you. Which is why it is harder to change than a rule — a rule can be memorised, a disposition can only be trained. It also explains a familiar embarrassment: you know perfectly well what to do, and your body is still half a beat late.
The standard objection to all this is a serious one: the concept risks circularity. Any behaviour whatsoever can be labelled habitus after the fact, which means it forbids nothing and cannot be falsified; and it is extremely hard to measure. State the evidence plainly: habitus is a concept grown out of ethnographic material, not a testable prediction. Its value lies in making visible that a batch of differences usually filed under "personal style" have a systematic source.
At a parents' evening a teacher says a child "expresses herself well". Which class taught that? There is no such line on the report card, yet it enters the assessment.
Bourdieu and Jean-Claude Passeron, in Les Héritiers (1964) and La Reproduction (1970), aimed at exactly that gap: French higher education admitted wildly unequal proportions by social origin while the school sincerely believed it was selecting fairly. Their claim is not that schools cheat. It is that part of what a school rewards is not in its curriculum — that part is picked up at home, tuition-free but far from cost-free (the cost being time, books on shelves, the way adults talk). Sociology calls this cultural capital. The "cultural fit" seen in elite hiring in the social stratification topic is the same machine at the workplace end.
Bourdieu split it into three states, and the split is more useful than the term. Embodied: accent, bearing, a feel for which topic is appropriate — only absorbed over years. Objectified: the books, instruments and pictures in a house — purchasable, though owning is not the same as being able to use. Institutionalised: the credential. That third step is the decisive leap: a diploma converts "my family had these things" into "I hold this qualification", and the origin is severed at the moment of exchange.
Can any of it be measured? Paul DiMaggio tried in the American Sociological Review in 1982, using eleventh-graders in the 1960 Project Talent survey and factor-analysing responses like wanting to attend classical concerts, reading serious literature, and thinking of oneself as cultivated. Net of ability test scores and family status, the measure still predicted school grades, with different effects for boys and girls. Evidence: one cross-sectional correlation on 1960 US data; replications elsewhere find the direction more often than not but magnitudes vary widely — this is not a firmly replicated law.
More vivid is Annette Lareau's Unequal Childhoods (2003). She and her assistants interviewed 88 children and their parents, and followed 12 families closely. One of the sharpest settings she records is the paediatrician's office: middle-class children, coached by their parents beforehand, ask the doctor questions and even correct him; working-class and poor children are taught to be quiet and respectful in front of authority. She names the two styles concerted cultivation and the accomplishment of natural growth.
One of her points is routinely dropped by people citing her: the second is better for children in many respects — far more free play, closer ties to relatives. Its disadvantage is not in the child; it is that institutions like schools and hospitals do not cash it. Evidence: small-sample ethnography, mechanism clearly visible, proportions not to be extrapolated.
Someone says a thing is "a bit tacky", or "trying too hard". You can more or less feel what they mean, and nobody can state the standard. Bourdieu's La Distinction (1979, translated 1984) is an entire book taking that feeling apart.
The data are ordinary survey data: two rounds, in 1963 and 1967–68, 1,217 people in all, asked about music, painting, photography, furniture, food and dress. Preferences map systematically onto educational capital and parental occupation. What grows in permanently straitened circumstances he calls the taste of necessity — substantial, sufficient, function first, a photograph should show what it is a photograph of — while at the high-education pole attention goes to form rather than content, and one can talk at length about a picture with "nothing in it".
His most counterintuitive claim is about mechanism: taste operates mainly by refusal. Aesthetic opposition is social opposition first — saying "too sweet", "too flashy" draws a line between yourself and someone. So ask a person what they like and the answer is often vague; ask what they cannot stand and their position is immediately legible. And which taste counts as legitimate is not a natural growth: it is underwritten by schools, museums and critics. The front-stage performance in the dramaturgy topic is a deliberate management of impressions; this layer is different — what strikes you as overdone is not something you perform, it is something you cannot help.
How much of this still holds? The most important revision comes from Richard Peterson and Roger Kern in 1996: comparing the 1982 and 1992 rounds of the American Survey of Public Participation in the Arts, they found high-status people no longer consuming only highbrow culture but ranging visibly wider, a change driven both by cohort replacement and by change within the same people. They call the type a cultural omnivore. This does not overturn distinction; it changes its form: the marker moves from what you listen to toward how many kinds and in what stance — listening to everything and speaking well about each has itself become the scarce thing.
The strongest counter-position deserves stating as its authors would. John Goldthorpe's 2007 critique makes two points: cultural capital is loosely formalised with a weak causal chain, and educational inequality can be explained without it at all. The relative risk aversion model he proposed with Richard Breen in 1997 assumes only one thing — that at every educational branching point, families of every class aim above all to avoid downward mobility, and since "downward" means a different distance from different starting points, that asymmetry alone generates class-differentiated choices. The two have been compared head to head: a 2007 study of Amsterdam secondary students concluded that they govern different links in the chain — cultural capital more closely tied to school performance, relative risk aversion to which track is chosen at equal performance. Evidence: a single-city sample, with results inconsistent across countries. The dispute has not converged; neither side should be taken as settled.