Dance is the only art whose instrument dies. That single fact explains almost everything about how it is made, taught, lost and fought over.

In 1935 a young American anthropologist and dancer named Katherine Dunham took fellowships from the Rosenwald and Guggenheim foundations and went to the Caribbean. She spent about eighteen months in Jamaica, Martinique, Trinidad and Haiti, and she did not go as a tourist collecting steps. She trained, she was initiated, and she wrote a thesis on the dances of Haiti, on their material aspect, organisation, form and function.

What she did with the material is the reason her name matters. She built a codified technique out of it, grounded, polyrhythmic, hips and spine mobile in ways ballet keeps still, and she ran a company on it for close to thirty years, at a time when it was the only self-supporting Black American dance company in the country. The Dunham Technique is still taught.

She kept the relationship for the rest of her life. She bought Habitation Leclerc outside Port-au-Prince. In 1992, aged 83, she went on a forty-seven-day hunger strike against the forced repatriation of Haitian refugees, and Haiti later gave her its highest honour. She died in 2006, aged 96.

Dunham is the argument of this article in one biography: that dance is knowledge carried in bodies, that it travels when people travel, and that treating it as decoration misses what it is doing.

The four things every dance is made of

Whatever the tradition, dance is organised along four axes, and knowing them turns watching into reading.

Body. What moves, and what stays still. Which parts are articulate and which are held. This is the fastest way to identify a tradition: ballet lifts and lengthens and keeps the pelvis quiet; a great deal of African and Afro-Caribbean dance is grounded and mobile through the spine, ribcage and hips; South Asian classical forms carry enormous detail in hands, eyes and neck.

Space. Where the body goes and what shape it makes. Level, direction, pathway, and whether the dance faces an audience at all. A dance made for a circle and a dance made for a proscenium stage are different objects, and putting the first inside the second changes it.

Time. Not just speed, but the relationship to the music. Dancing on the beat, behind it, or against it are three different aesthetics, and a dancer who is never quite on the beat may be executing the style precisely.

Energy. How movement is done: sudden or sustained, direct or indirect, heavy or light, bound or free. Two dancers can perform identical steps and produce entirely different dances, and this is the axis that difference lives on.

The notation problem, and why dance keeps disappearing

Music has had a workable notation for centuries. Dance does not, in any comparable sense.

Systems exist. One was published in the 1920s and another in the 1950s, both capable of recording movement in real detail, and both requiring specialist training to write or read. Neither is in general use. The overwhelming majority of the world’s dance is transmitted the way it always was, from one body to another, in a room.

Two consequences follow, and they are large.

The first is that dance is the most fragile art form there is. When a generation of practitioners dies without teaching, the repertoire goes with them. There is no equivalent of a score sitting in a library waiting to be played again. Film helps and is not the same thing: a recording shows you what a dance looked like from one angle, not how it is done.

The second is that dance is inseparable from lineage. A dancer can usually tell you who taught them, and who taught their teacher. That chain is the archive, and it is why questions of who is authorised to teach a form are so much sharper in dance than in painting.

Ballet is a regional folk form with very good branding

Ballet’s vocabulary is French because the form was codified at the French court. Louis the Fourteenth, himself a dancer, established a royal academy of dance in 1661, and out of that institutionalisation came the positions, the turnout, the naming and the teaching method that professional ballet still uses.

This is worth saying plainly, not to diminish ballet, which is an extraordinary technical achievement, but because it is routinely treated as dance itself rather than as one tradition among many. A form that developed in one country over a few centuries, for a court audience, became the standard against which other dance was measured, and a great deal of nonsense about which dances are sophisticated follows from that.

The honest comparison runs the other way as well. By the standards of a Haitian Rada dance, ballet is rhythmically simple, because the body is doing one thing at a time against one pulse. By the standards of ballet, the same dance is spatially simple. Each is demanding in the dimension the other neglects.

When dance is not expression

The assumption built into most Western writing about dance is that it expresses an inner state: the dancer feels something, and the movement conveys it.

In a great deal of the world’s dance this is simply not what is happening.

In Haitian Vodou, dance is closer to a technology than to an expression. Specific dances belong to specific lwa, in the way the drum rhythms do. The body performs a form associated with a particular spirit, and the purpose is arrival: at the point of possession, the lwa is understood to be present and the person is not performing them. What the dance conveys is not the dancer’s feelings. The dancer is the site.

That distinction reframes everything about how such a dance should be watched and, more practically, about what a staged version of it is. A Vodou dance in a theatre, performed by trained dancers to an audience with tickets, is a real and legitimate art object, and it is not the thing it is derived from. Dunham was unusually clear about this, which is why her work holds up: she made concert dance out of the material without claiming it was the ceremony.

Dance as a claim on public space

Because dance needs bodies and room, it is a public act, and governments have paid it a great deal of hostile attention.

The pattern is consistent across centuries and continents: dances associated with enslaved and colonised people are banned or licensed, on grounds of public order or morality; the dances survive anyway, often by being carried into religious observance or festival, where they are harder to prohibit; and two or three generations later they become the national culture, taught in schools and sold to visitors.

Carnival is the clearest example of the mechanism. A sanctioned period in which the street belongs to the crowd, the ordinary rules are suspended, and dance takes over the public space of the city. Rara in Haiti does something similar in the Lenten season, moving through the streets with its own music, its own hierarchy and its own right of way.

Social dance has its own version of the same story. Nearly every popular partner and street dance of the last century came out of a Black or working-class community, was described as a danger to the young, and became respectable within about twenty years. The steps change. The sequence does not.

The families of dance, and what the usual map leaves out

Lists of dance styles tend to be alphabetical, which tells you nothing. Sorted by what the dance is for, the field becomes legible.

Court and classical forms were made for rulers and codified by institutions. Ballet, the classical dances of South and Southeast Asia, the court dances of Korea and Japan. They share long training, a fixed vocabulary, and an authority somewhere deciding what is correct.

Social and partner dance exists so that people can do it together without training. Its evolution is fast, its transmission is informal, and its rules are social rather than aesthetic. Most of the twentieth century’s partner dances came out of the Caribbean and the Americas, and the ballroom versions taught in studios are standardised descendants, not the source.

Street and vernacular dance is made by communities and judged by them, frequently in competition. Its distinguishing feature is that the standard is set by peers in a circle rather than by an academy, which makes it the fastest-evolving category and the one most thoroughly appropriated.

Ritual and ceremonial dance exists to accomplish something rather than to be watched. Attendance is participation, not spectatorship.

Concert dance is the modern category: dance made for a paying audience in a theatre, drawing on any of the above. It is barely a century old in most of the world and it is where the other four meet, not always with permission.

What the standard map leaves out is the traffic. Nothing in this list is sealed. The technique Katherine Dunham built is ritual material turned into concert vocabulary; ballroom Latin is social dance turned into a competitive sport; and a great deal of contemporary choreography is street vocabulary taken into the theatre. The interesting questions are always at the borders: what was kept, what was dropped, and whether the people who made the material are in the room.

The working life, which nobody describes

Dance is the art form with the shortest professional career and the least money in it, and the two facts are related.

A classical dancer’s performing career typically ends in their late thirties, and ends earlier if an injury arrives at the wrong moment. Training starts in childhood for classical forms, which means the investment is made by families long before anybody knows whether it will pay. Company contracts are seasonal, freelance work is the norm, and rehearsal is frequently the unpaid part.

Injury is not an accident in this profession; it is an occupational certainty that companies manage with varying honesty. And the physical demands do not pause for the economics: a dancer must train daily to stay employable, whether or not they are being paid that week.

This is worth knowing before watching, because it changes what you are looking at. A company of twenty on stage represents an enormous quantity of unpaid preparation and a narrow window in which it can be spent. It is also why teaching matters so much in dance: the second career is built into the first, and the lineage that keeps the art alive is maintained by people whose performing years are over.

What contemporary dance is actually doing

Modern and contemporary dance are often described as freedom from ballet, which is half the story and the less interesting half.

What the twentieth century’s choreographers really did was question the assumptions underneath ballet one at a time. That dance should be up rather than down. That it should be beautiful. That it should tell a story. That it needs music. That it needs trained dancers. That there should be an audience on one side and performers on the other. Each of those was taken away by someone to see what remained.

That is why contemporary work can look difficult. It is frequently an argument about the medium conducted in the medium, and if you do not know which assumption is being removed, you are watching the answer to a question you did not hear.

The useful question in front of any contemporary piece is therefore not what does it mean but what is it refusing. Gravity, prettiness, narrative, the fourth wall, virtuosity itself. Identify the refusal and the piece usually opens.

How to watch dance

Watch the feet, then watch the spine. Most audiences watch the arms and the face. Weight and spine tell you the tradition and the quality.

Find the relationship to the beat. On, behind, against, or indifferent to it. This alone distinguishes most styles.

Notice the floor. Is the dancer resisting gravity or using it. Ballet goes up. A great deal of the world’s dance goes down, and the ground is a partner rather than an obstacle.

Watch one dancer in a group piece for a whole section. Ensembles are designed to be seen as a mass, and following one body shows you how much individual labour the mass is made of.

Notice where you are sitting. A dance made in a circle and shown on a stage has been rotated for you, and something has been lost in the rotation. Knowing that is part of watching honestly.

Where to start, specifically

Read Katherine Dunham. Her writing on the dances of Haiti is the rare case of a major artist who was also a trained anthropologist writing about the source of her own work, and it models the relationship between research and practice that this magazine keeps arguing for.

Then find a Dunham Technique class if you can, or a video of one. Ten minutes of watching the exercises shows you the grammar her company was built on, and you will recognise it immediately in a great deal of American concert dance that never credits it.

In New Jersey, look at what is programmed at the New Jersey Performing Arts Center in Newark and at the university dance departments, which present work that never tours commercially. And when a company comes through with a piece drawn from a ritual tradition, read the programme note carefully. A company that tells you who taught them and where the material came from is telling you something about the integrity of the whole production.

And the exercise that costs nothing: put on music you know well and try to move only your ribcage, then only your hips, then both against each other. Five minutes of that will change what you see the next time you watch someone who can do it.

This article consolidates and replaces seven shorter pieces previously published separately on dance as universal expression, the elements of dance, dance styles around the world, traditional dance and cultural heritage, the role of dance in ritual and celebration, contemporary dance, and dance as identity and social expression. The earlier URLs now redirect here.

Sources for dates and attributions: biographical records of Katherine Dunham, 1909 to 2006, on her Rosenwald and Guggenheim fellowships of 1935, her eighteen months of fieldwork in Jamaica, Martinique, Trinidad and Haiti, her thesis on the dances of Haiti, the Dunham Technique, her company, her purchase of Habitation Leclerc, and her forty-seven-day hunger strike in 1992 against the repatriation of Haitian refugees; standard histories of the French royal academy of dance established in 1661.