Ritual as Music – on Manuel de Falla's El amor brujo

Pastora Imperio, the Andalusian flamenco artist who commissioned El amor brujo in 1915. Public domain · Wikimedia Commons

I. What the Fire Is Part Of

It is rare for so short a scene, so short a piece of music, to have taken on a life of its own to the extent that Danza ritual del fuego has. Five minutes of orchestral music, originally composed as a single dance number within a larger stage work from 1915, has since become one of the most-performed individual pieces in the twentieth-century repertoire. It has been transcribed for piano, inserted into films, used as an audible symbol of the "Spanish" in contexts that have nothing to do with El amor brujo. Concert audiences often "know" it before they know what it is.

As I now prepare to conduct El amor brujo, I see that part of my work is precisely to resist this – what I see as a populist side to the fire dance. Not by playing the dance down when it comes – it must have all its power – but by remembering, throughout, that it is not a kind of crowd-pleaser, but a scene. The fire has been lit so that Candela can burn something out, drive something away. It is my view that everything that happens before that scene builds up towards it, and that everything that happens afterwards builds down from it. The work is a ritual in music, and the dance is the ritual's middle, its centre.

That is the angle from which I want to write here. El amor brujo is not first and foremost a work about Spanish folklore lifted into a modern orchestra, even though it is also that. It is a work in which a dramatic action – an exorcism – is carried by the music itself. The music does not depict the ritual from the outside; the music is the ritual, it performs it. That changes what has to happen in the rehearsal room, and what has to be heard in the hall on the night of the concert.

II. From Pastora Imperio to 1925

El amor brujo has a genesis that is unusually long for so short a work. Manuel de Falla composed the original version in 1915, on a commission from the great Andalusian Gitano artist Pastora Imperio – one of the foremost flamenco dancers and singers of her time, whose family also contributed textual material drawn from the Gitano tradition. The premiere took place in Madrid in April 1915. Its reception was cool, partly negative; the work was a hybrid between theatre and concert that the Spanish public apparently did not quite know what to do with.

De Falla did not give up. He revised the work several times in the years that followed. In 1916 he made a concert version for larger orchestra. In 1925 he completed the ballet version for mezzo-soprano and full orchestra which has ever since been the most-performed form. It is this 1925 version I myself am now sitting with, working on and preparing to conduct, and which most recordings and concert performances today refer to.

The background for de Falla's being able to write such a work at all lies in his own biography and in the aesthetic position he had arrived at in the years before the war. He had lived in Paris from 1907 to 1914, where he was in close contact with Debussy and studied with, among others, Paul Dukas. He returned home to Spain shortly before the First World War, and eventually settled in Granada – near the Alhambra, near the Gitano quarter of Sacromonte, near the Andalusian milieu that would become the raw material for much of his subsequent composition.

By then he had a clear aesthetic agenda. Spanish music had over the course of the nineteenth century been largely defined from the outside – by the Spanish exoticism of Italian operas, by Bizet's Carmen (a French composer's fantasy of Andalusia), by the Spanish-tinted character pieces of salon pianists. Manuel de Falla believed that Spanish music had to rediscover its own core in the living folk-musical traditions, and above all in the flamenco milieu of Andalusia, and that a modern composer quite simply had a duty to transform these traditions into concert music without falsifying them.

The material of El amor brujo is Gitano-Andalusian throughout. Candela, a young Gitano woman, is haunted by the ghost of her former lover – a man who had treated her badly, and who now, from the other side, prevents her from loving another man, Carmelo, who wants her. An older woman from the community suggests that there is a way to break the possession: an expulsion through fire and dance, together with a ruse in which another woman, Lucía, lures the ghost away so that Candela and Carmelo can finally kiss each other and thereby loosen the old grip.

The purification ritual takes place in the fire dance. As Candela dances around the fire, she tries to burn the ghost out of her own body. Every rotation, every stamp, every movement is part of the action. It is not a dance about something; it is a dance that does something.

It is this material de Falla takes and writes into an orchestral landscape that is at once authentic and advanced. He uses Andalusian modes – the Phrygian mode with its characteristic lowered second, melismatic melodic lines – and he combines them with dissonant harmonies, refined orchestration and a formal structure that has more in common with Debussy and early modernism than with flamenco. The result is a work no Gitano would have written, but which no composer from outside could have written either without deep respect for the material he was entering into.

III. Composing an Action

The way I read El amor brujo – and this is a conductor's reading, not a musicologist's verdict – its great achievement is not first and foremost the transformation of folklore, but the fact that an entire orchestral work is structured according to the logic of a ritual.

A ritual has its own form. It opens with a call that gathers attention. It names what is to be dealt with. It builds towards a culmination in which something actually happens – something is driven out, something is cleansed, something is transformed. And it closes by leading the participants back to the everyday world, often through some sign that the night is over and the day has returned.

El amor brujo follows this form precisely. The opening is a call, hard and modal, that immediately establishes that we are not in a concert hall, but in a room where something is going to happen. Candela's first song, Canción del amor dolido ("Song of Wounded Love"), names the pain. The intervening dances and scenes describe the attempts to deal with it. The fire dance is the culmination itself – the action in which the possession is to be burnt out. And the final scenes, with the love play between Candela and Carmelo and the morning bells ringing in the dawn, lead us back to the everyday world.

For me as a conductor, this has practical consequences that are quite specific to this work. I quite simply cannot treat El amor brujo as a suite of Spanish dance numbers with a famous third movement. I have to hold the ritual's dramaturgical arc intact, even when I conduct the concert version without dancers on the stage. That means it is absolutely necessary to know, at every moment, where in the action we are. Are we calling in the attention? Are we naming the pain? Are we approaching the fire? Is the fire under way? Is it burning down? Have we led Candela and Carmelo safely through the night?

That is the awareness which, in my view, must lie beneath the interpretation. Without it, the individual movements become splendid Spanish orchestral pieces, but the work as a whole disappears. With it, even the smallest transitions become meaningful, because they are not merely musical transitions – they are phases in an action.

This also means something concrete, for me, about how the individual sections should be played. I believe, for example, that the opening should not be grand and effectful; it should be gathering – like a clap of the hands asking for silence before the ceremony begins. Candela's song is not an operatic aria; it is a lament. The fire dance is not virtuosity; it is work. And the ending is not triumph; it is a daybreak that acknowledges the night has been long.

That is the reading I bring with me when I open the score. It is less about "Spanish versus modern" and more about "ritual action carried in the form of music".

IV. Three Places in the Ritual

Three places in the score are especially decisive for me – three places where the logic of the ritual is either held or lost.

The first is the opening and Candela's first song. El amor brujo does not open gently. The first thing we hear is a hard, rhythmic figure in the orchestra – woodwinds and strings together in an aggressive, modal gesture that immediately places us in an Andalusian sound-world. It is not a quiet introduction; it is a call, like someone striking the table to ask for attention before something important is to be said.

After this call comes Canción del amor dolido. The mezzo-soprano sings a long, melismatic line that lies close to flamenco's cante jondo tradition – deep song, modal, often with microtonal inflections that Western notation cannot quite capture. De Falla writes it down in standard notation, but what is required of the singer is a highly specific vocal style that lies close to flamenco practice.

For me as a conductor this is one of the most difficult places in the whole work, precisely because it looks simple. If the opening is too fortissimo or too theatrical, we have already slipped into an exotic-effect interpretation before the work has begun. If Candela's song is not given the acoustic space it needs, the weight of the lament disappears. What I have to do is hold the opening at a controlled strength – a gathering call, not a roar – and then draw the orchestra back so that the singer has enough air to fill the room without having to force. It is one of those gestures in the score that decides whether the whole remainder of the work can succeed: if the ritual's first two steps – the call and the naming – do not sit, the audience does not enter the action.

The second is, inevitably, Danza ritual del fuego. The fire dance comes in the middle of the work, as the dramaturgical turning point. Candela has resolved to drive the ghost out, and she does so through the dance around the fire.

De Falla builds the dance as a long, controlled acceleration. It opens with a soft, repeating figure – a whipping motion in the strings, a heartbeat pulse in the timpani. Over this he places a melodic line (first in the oboe, later in solo violin and flute) that spins like a spiral figure – the melody goes round and round itself, in the same way as the dance physically spins around the fire.

The acceleration is gradual. The tempo should not lurch forward; it should take hold, as fire itself takes hold. The orchestra fills out gradually – more winds enter, more percussion, until finally the whole orchestra is at fortissimo. The climax arrives as a crash. And then – this is the decisive point – the fire burns itself out. The dance ends not in triumph, but in a burnt-out silence, with only a single lament left in the strings.

It is here, at this ending of the dance, that it shows whether I as conductor have held the ritual intact. If the audience has experienced the dance as a crowd-pleaser, they applaud – and the whole rest of the work collapses, because the action does not receive its release. If I have managed to make it clear, throughout the dance, that the fire is burning something – that this is work, not virtuosity – then the audience sits in silence afterwards, and the work can carry itself forward.

It depends on small things. On the acceleration being so gradual that we almost do not notice it happening. On the dynamic build following the tempo without being forced. On the sparks of fire in the orchestra that lie beneath the dance not being overplayed for effect, but held back until they find their proper place in the action. On the final dynamic drop – the transition from climax to burnt-out silence – not becoming a Romantic decrescendo, but an abrupt end: the fire has done its work, it goes out.

The third is the ending of the work. After the fire dance has done its work – or after Lucía's ruse has lured the ghost away – Candela can finally kiss Carmelo. This is played out in a light, playful, dance-like sequence: the dance of the love play. It has an entirely different character from everything that has gone before. It is daylight, lightness, a musical exhalation after a long night.

And finally come the morning bells, Las campanas del amanecer. The bells are rung in from a distance (de Falla specifically calls for this in the score). The orchestra closes the work with a short, uplifted chord that is almost Catholic in its purity – a gesture that sets the whole Gitano-Andalusian ritual within the larger, church-bell-marked world of which Andalusia is also a part. It is an ending that acknowledges we have been through something extraordinary in the night, and that we are now back in the light of day.

For me it is a matter of holding this ending calm and uplifted without making it sentimental. It is a gesture of a regained order, not of triumph. The ritual has succeeded; what was to be burnt has been burnt. Now everyone can go home.

V. Coda

El amor brujo is a work that is easier to underestimate than to hear rightly. It is short (around twenty-five minutes), it contains one world-famous dance, and it can easily be reduced to a Spanish-exotic concert filler with a barnstorming middle piece. It deserves more.

What, to my mind, makes the work special is not that Spanish folklore has been lifted into a modern orchestra (de Falla does that elsewhere too, not least in his other great ballet work), but that an entire orchestral work has been composed as a ritual – with a ritual's form, with a ritual's purpose, with a ritual's demand that something actually happen while the music is playing. It is a rare gesture in Western concert music. We have many works that depict rituals from the outside. We have fewer that try to perform them.

For me as a conductor, this means that I see the task here as somewhat different from what it is with most other works. It is not first of all to demonstrate the qualities of the score to the audience. It is to lead the audience through an action – to carry them into the call, through the naming of the lament, towards the fire, through the fire, out of the fire and back to the dawn. If I succeed in this, the audience will not sit and applaud the fire dance in the middle. They will sit and listen, and they will, one hopes, wait for what comes after.

As I now stand before the work again and am soon to perform it, it is for this task I am preparing myself. It is the arc of the ritual I hold in my hands. The opening's call, the lament's space, the fire's work, the morning bells' regained order – all of it rests on the same thing: that the music is allowed to be an action, not merely an interpretation of an action. If that is allowed to happen, de Falla has given us something unusually rare. He has given us a ritual we can go through together.