Life becomes more intense, the loves and hates of other worlds pass before our eyes, and we feel whatever is highest and lowest in ourselves.’ These words, recording the effect created by the flamenco dancer Pastora Imperio, evoke the potently suggestive world of Spain’s native art form, and its ability to unlock deep-seated wells of primal impulse and emotion.
How Manuel de Falla came to compose his ballet El Amor Brujo
In 1914, Imperio forged an unlikely link with a composer whose classical training appeared to place him well beyond the boundaries of the popular flamenco tradition: Manuel de Falla. Would he, she wondered, be interested in creating a new work for her to sing and dance in?
De Falla was, it turned out, more than merely interested. Like many, he was already in thrall to Imperio’s spell-binding artistry, and intent on using the rich folk music heritage of his country to inspire a new, distinctive national classical style. ‘It has occasionally been asserted that we have no traditions,’ de Falla wrote of Spain. ‘But in our dance and our rhythm we possess the strongest traditions that none can obliterate.’
Who was Manuel de Falla?
Born in Cadiz in 1876, Manuel de Falla forged his early career in Madrid. It was, however, the culture of his Andalusian birthplace which above all would infuse the works that made his name. A seven-year stay in Paris from 1907 saw him mix with the likes of Ravel, Debussy and Stravinsky plus impresario Serge Diaghilev, who would later commission El sombrero de tres picos for the Ballets Russes. For the majority of the 1920s and ’30s he lived in Granada but, following Franco’s victory in the Spanish Civil War, he took the decision to move to Argentina. He died there, aged 69, in 1946.
That is the spirit in which the original version of de Falla’s one-act ballet El Amor Brujo (‘Love, the Magician’) was created. Based on songs Pastora Imperio and her mother had sung to de Falla, and folk tales they had told him, the first version of El Amor Brujo was cast as a ‘gitanería’ (‘gypsy entertainment’), with songs, dances and spoken dialogue.
The libretto, largely by the writer María Martínez Sierra, centred on the efforts of a woman, Candelas, to cast off the baleful influence of her deceased husband’s ghost and marry a new lover. De Falla’s music, scored for a small chamber ensemble, was all newly written, though closely modelled on the ‘cante jondo’ (‘deep song’) style he had heard from the Imperio women.
The premiere of this initial version of El Amor Brujo was in Madrid on 15 April 1915, and de Falla’s hopes for it were high. ‘I have tried to “live” it as a gypsy, to feel it deeply,’ he said a few hours before curtain-up. ‘And I have used in it no other elements than those which I believed to express the soul of that race.’
But although both Imperio and members of her family were in the cast, El Amor Brujo Mark I was not successful. ‘The gypsies on the stage felt the music to be truly their own, and were enthralled,’ a friend of de Falla’s later reported. But ‘no one liked it, not the general public, not the intellectuals, and not the critics’, who, ironically, accused the music of lacking ‘Spanish character’.
De Falla was, however, convinced that there was merit in the piece, and soon began re-working it in search of broader audience approval. This involved ditching the spoken dialogue, scaling up the orchestration and cutting the number of songs. A revised version of El Amor Brujo was given a year later, but it took eight more years before the score we usually hear today was finally completed.
By then, El Amor Brujo had morphed into a half-hour ballet (a ‘ballet pantomímico’, de Falla called it), the Roma dance elements of the original translated to a more mainstream style of choreography. The flamenco vocal writing of the 1915 original had been distilled to three set-piece songs for mezzo-soprano – though some modern recordings still use an authentic flamenco cantaora for these – and the orchestra expanded to standard classical dimensions.
Why did de Falla make these wholesale changes? Some argue they dilute the visceral impact of the original ‘gitanería’, sanitising the raw emotions it contains, with the supernatural threat of an abusive former partner ever-present. Several fine recordings of the ‘gitanería’ have been made, and they to some extent confirm it as the rawer, more disturbing experience.
But de Falla knew the 1915 version was ultimately destined to have niche appeal only, and felt that the rich, deep-rooted sounds of Andalusian music he had captured in his score deserved a bigger platform. This it finally received on 22 May 1925 at the Théâtre du Trianon-Lyrique in Paris, when the definitive El Amor Brujo was premiered, with de Falla himself conducting.
The evening was, as one commentator puts it, ‘a roaring success’, and several movements from de Falla’s brilliantly evocative score – the ‘Danza ritual del fuego’ (‘Ritual Fire Dance’), in particular – were soon being performed in the concert hall. The critics too, it seemed, were finally happy. ‘Purity of line in the writing,’ composer Charles Koechlin wrote, ‘simplicity amid richness, and unexaggerated originality simply leap forth from this work.’
El Amor Brujo remains globally popular today, and stands as a striking example of how arguably Spain’s greatest composer deftly married the age-old ethnic traditions of his native country with the more formal parameters of the classical music tradition.
The best recordings of Manuel de Falla's ballet El Amor Brujo
Pablo Heras-Casado (conductor)
Marina Heredia (singer); Mahler Chamber Orch.
Harmonia Mundi HMM902271
This 2019 recording was the Mahler Chamber Orchestra’s first of Spanish repertoire, but their exceptionally vibrant performance belies their relative inexperience of the idiom. Conductor Pablo Heras-Casado is a major factor in the success of the interpretation. Spanish himself, and like de Falla an Andalusian, he has this music in his bones and communicates its swirling passions to the orchestra with total conviction.
Heras-Casado has a particularly acute ear for woodwind detail. The flute, piccolo and oboe lines lend an appealing edge of sharpness in the brief ‘Introduction’, while the solo clarinet in the gypsy cave has a pleasingly guttural quality, and the oboe fibrillates sensually in the lead-in to the ‘Song of Suffering Love’.
The songs bring one potential point of controversy, as Heras-Casado elects to use the flamenco singer Marina Heredia as the soloist, not the classical mezzo-soprano found in most other recordings. The effect is initially startling, but the combination of Heredia’s smoky-toned, intense vocalism and the slicing string staccatos of the ‘Song of Suffering Love’ make a rivetingly authentic impression.
‘The Apparition’ has a skirling momentum, the muted trumpet spitting fire and the piano tracing weird shapes in the ether. Here, and in the ‘Dance of Terror’ which follows, Heras-Casado skilfully marries a wealth of instrumental detail with a mounting sense of fright and trepidation.
In the famous ‘Ritual Fire Dance’, where Candelas attempts to exorcise the ghost of her dead husband, Heras-Casado wisely respects de Falla’s request to avoid a too-fast tempo, while nonethess conjuring a thrumming sense of menace. There’s insinuating solo work from the oboe and stirring contributions from both the horn and violin departments. The Mahler Chamber Orchestra has no weak links – it is packed full of technically excellent and expressive players.
More than any other conductor, Heras-Casado taps into the ‘gitanería’ element of El Amor Brujo, the chamber-music clarity of the textures he favours – in the ravishing ‘Pantomime’, for instance – clearly inspired by the small-ensemble scoring of the original 1915 version. The recorded sound is excellent, and a scintillating performance of El sombrero de tres picos (‘The three-cornered hat’), de Falla’s other great ballet, makes the album as a whole an irresistible proposition.
Charles Dutoit (conductor)
Decca 410 0082
Charles Dutoit’s recording of El Amor Brujo was made in 1981, just as his golden period with the Orchestre Symphonique de Montréal was beginning. It’s a typically virile reading, more conventionally ‘symphonic’ in sound than Heras-Casado’s leaner version, with a big, transparent soundstage framed by Decca’s expert engineering team. Occasionally, the performance seems a touch overdriven and rhythmically dogmatic. But it benefits from the spirited contribution of Canadian mezzo-soprano Huguette Tourangeau, who injects a welcome element of earthiness into her vocal delivery.
Igor Markevitch (conductor)
Philips 484 2777
The Ukrainian Igor Markevitch, a marvellous conductor, made his recording with the Spanish Radio and Television Symphony Orchestra in 1966.
It’s a performance which deserves to be much better known, with a tensely atmospheric ‘In the Cave’ and a tingling ‘Dance of Terror’. Inés Rivadeneira is a ripe-toned presence in the songs, adding strong flamenco inflections to her solidly schooled classical mezzo. A well-balanced analogue recording catches the incisive musicality of Markevitch’s interpretation in a highly satisfying fashion.
Rafael Frühbeck de Burgos (conductor)
Decca 466 1282
Like Heras-Casado, Frühbeck de Burgos uses an authentic flamenco artist – the smouldering Nati Mistral – for the songs, adding spice and bite to his vivid, sharply rhythmic take. The playing of the mid-1960s New Philharmonia is strongly characterful, with exotic oboe solos and swashbuckling string playing in the ‘Ritual Fire Dance’, and seductive phrasing by the conductor. The Decca engineers secure a classic analogue recording – rich, beautifully balanced and sensitively attuned to the ballet’s many variations of tint and timbre.
And one to avoid…
Ernest Ansermet was a famed conductor of ballet, but his 1955 recording of El Amor Brujo with L’Orchestre de la Suisse Romande is curiously flat-footed. The ‘Introduction’ drags its heels significantly, and the ‘Dance of Terror’ also suffers from a lack of propulsion. The opening of ‘Pantomime’ is similarly laboured, and mezzo Marina de Gabaráin is scrunched to the far left of the stereo spectrum. The results are, in short, underwhelming.