The arts were loosely defined by a plethora of ‘-isms’ in the second half of the nineteenth and the early twentieth centuries. None is more often associated with Debussy than Impressionism. Even recent scholarship is still disposed to position him as an Impressionist composer. Whilst much work has been done to disentangle Debussy from the tag and align him in relation to, among others, Hellenistic paintings (around the time of the Prélude à l’Après-midi d’un faune [it.]), Symbolist painting, and the English Pre-Raphaelites, it is important to understand what has been intended by the term ‘musical Impressionism’, how it came to be associated with Debussy, and his usually hostile response to being thus categorised.

However perverse it might seem to begin the story of Carmen with discussion of a ruddy-faced British battle-commander, the Duke of Wellington was a key figure in the chain of events preparing the ground for Bizet’s celebrated opera long before it was first staged in 1875. The influence on the course of European history of the Irish-born Arthur Wellesley was tremendous. Most celebrated for his victory over Napoleon at Waterloo in 1815, for the Spanish – and for an important Spanish strand in the development of French music – his earlier victory over Joseph Napoleon (Bonaparte’s brother) in 1813 at the battle of Vitoria in northern Spain was crucially important. It was decisive in Spanish history, effectively ending Napoleonic rule and the Peninsular Wars in which Spain, England and Portugal had allied to oppose the invading French armies. In that battle, on 21 June 1813, Wellington finally vanquished the Napoleonic forces, and restored the Spanish crown to Fernando VII. The musician Narciso Paz dedicated his Troisième collection d’airs espagnols to Wellington in recognition of his liberation of Spain, though the collection was published in France.

By the time Wellesley entered the fray in the Iberian peninsula the British armies there had recruited and trained a substantial number of Portuguese troops and several Spanish generals commanded trustworthy battalions, most notably General Álava. One of Wellington’s strengths was in his meticulous organisation of a supply chain, which was vital if an army was to survive the harsh extremes of weather in the north of Spain. By contrast, the French armies had less back-up: writers on the period often quote a saying that ‘In Spain, large armies starve, and small ones get beaten.’ Thus, although the majority of paintings of the decisive battle of Vitoria were by British artists, it is a mistake to assume that the outcome was a triumphalist victory for the British. Nevertheless, on his return to England Wellesley was hailed as a hero, and it was his victories in the peninsula which earned him his dukedom, after already having been honoured by the Spanish. For Spain, the battle of Vitoria was a decisive event, resulting in the end of the era Josefina (1808–1813).

Juxtaposing the opera with the Mérimée sources has so far compared novella and libretto and their transformation into opera. It has also identified a few possible sources of Bizet’s ‘inspiration’, but do these really get to the heart of the characters of this spectacle? Counteracting the accusations that Bizet’s opera slightens the novella, a reverse approach stemming from French nineteenth-century gypsophiles – amongst whom Mérimée was one of the most committed – reveals the seriousness with which the two librettists approached their commission, transferring the essence, spirit and locations into a text ripe with ‘realism’ and a visceral presentation of this troubled tale. Halévy’s notebooks reveal him combing the first published collection of Mérimée’s correspondence, which appeared exactly in the year he and Meil-hac were transforming Carmen into a libretto.Above all, his notes reveal a search for an understanding of Mérimée’s deep attachment to Spain. Examination of both the first run and Albert Carré’s ‘remake’ of 1898 further unveil considerable team-efforts to capture these essential Spanish qualities – Don Preciso’s ‘El chiste y la sal de España’ – both on the stage and in the pit.

Frowned upon by some modern theatre analysts is the old-fashioned and dubious method of the ‘character sketch’, which was the backbone of school-level study. This has been attempted many times in relation to Carmen, but perhaps without much regard to Carmen’s foils, Micaëla, José and Escamillo,

Not until the 1860s was it possible to go from Paris to Spain entirely by train. French railways got near – to Bayonne – by 1856, but it was not until 1864 that the Spanish linked up their railway to Irun, still to this day the frontier junction. After the peninsula wars, the Spanish government had ensured that no French trains could cross into Spain by introducing a wide-gauge permanent way of six Castillian feet, just in case another Napoleonic-style invader harboured any ideas of steaming through the border. Despite this, Spanish theatrical companies (rather than individuals) began to get to Paris by the 1850s and grace its stages. How troublesome it must have been to go by boat, as was previously the way.

Spain was by this time opening up. Import and export became easier. In terms of musical spectacle Spain had become too much of an importer: of Italian opera and French ballet, as Don Preciso had complained. As the century progressed things began to change and, above all, Madrid and Barcelona began to realise they no longer had to rely on imported spectacle; they too could export culture. Opera was out of the question – the Italians and French could hardly be rivalled. But dance became a possibility, and companies were established in these theatres with French dancers prominent in the troupes. The Teatro del Liceu opened in Barcelona in 1847, closely followed by Madrid’s Teatro Real in 1850. A fusion of Spanish folk-dances and ballet seemed to be the way forward, and travelling companies were inaugurated, largely under French control, thus forming links with Paris.

The different dance traditions emerging in Spain are often confused. Stemming from the folk dances of the seventeenth and eighteenth centuries was a hybridisation with classical ballet, which professionalised these dances for the stage – they became a spectator performance danced no longer in soft slippers, but rather in proper ballet shoes. Based on what we now call the bolero, the style was known as the Escuela bolera (Bolero School), and it underpinned Spanish dance in the mid-nineteenth century, achieving considerable foreign acclaim. The Spanish were known above all for their dance.

In France ‘going Spanish’ for a couple of ballets became highly fashionable.

The plethora of documents pertinent to the first runs of Carmen both inform and confuse: no single document captured what it looked or sounded like in its first creation. Yet there are many rich sources that elucidate each other in a number of ways. Only by considering them together – and in the context of the Opéra-Comique at the date of Carmen’s premiere in 1875 – do we begin to get a sense of the importance of Bizet’s opera – particularly its role as the seed-bed of French realism.

The path to the stage had begun in June 1872, when Bizet had announced to the composer Paul Lacombe that the Opéra-Comique had commissioned a new work from him, Meilhac and Halévy. It was to be ‘une chose gaie’ (we might say a ‘happy piece’) and ‘aussi serré que possible’ (‘as succinct as possible’).This may well have referred to the tendency of some opéra-comique plots to become too complex and convoluted. The various hurdles Bizet and his librettists encountered on the rocky road to the stage of the Opéra-Comique, especially his brushes with the directorate there, have been many times recounted and will not be revisited here.

For a new opéra-comique the essence of the process of publication was that the text and score were always published separately. The livret (libretto), with both the spoken and sung text would be produced by a literary publisher, usually in a small format. The musical score would be handled by an established music publisher (in Carmen’s case, Choudens) accustomed to producing and distributing orchestral matériel. The vocal score did not include the spoken dialogue, but only répliques (cues) from the text to indicate when each musical number should start.

There were two different printings of the libretto published in the year of the first production:

1. • Henri Meilhac et Ludovic Halévy: Carmen. Livret. Opéra-comique en 4 actes. Tiré de la nouvelle de Prosper Mérimée. 8o. 87pp. Paris, Calmann-Lévy. n.d.

2. • Henri Meilhac et Ludovic Halévy: Carmen. [Livret] Opéra-comique en quatre actes. Tiré de la nouvelle de Prosper Mérimée. 12o. 68pp. Paris, Michel Lévy frères. Paris, 1875.

It was also included in Meilhac et Halévy, Théâtre, vol. VII, in a new printing.There are only minor disparities between these three sources.

One of the principal differences between the first Opéra-Comique vocal score of Carmen and subsequent editions was that it contained a soon-abandoned ‘Scène et Pantomime’ placed after the opening chorus, which disappeared from the stage around the time of Bizet’s death. It appeared in the copied conducting-score used at the Opéra-Comique and in the hand-copied orchestral parts used in performances until printed parts were produced in the 1890s.

Many commentators have written about it, but only a few recent editions have included it either within their main text or as an appendix.Most subsequent writers, from Pigot in his Bizet of 1886, to Macdonald in his biography of 2014, have expressed the view that cutting it improves the opera, largely without considering the case for its inclusion.Important biographers in between these gave it short shrift: Mina Curtiss mentions it as a ‘a typical little opéra-bouffe pantomime in which a young woman on the arm of her elderly husband flirts with her lover as they cross la place while the brigadier, Moralès, describes the episode in song’.Yet it certainly was not ‘typical’, and nor does Moralès really sing much! Malherbe and Winton Dean briefly mention it, but Streicher, Maingueneau and Lacombe fail to give it time of day. Most influential in its subsequent demise was probably Pigot in his first biography of Bizet.

Pigot had not changed his mind about the scene in his revised and substantially altered edition of 1911. Quite the reverse, he expanded his critical section on the scene, possibly influencing twentieth-century producers of the opera by suggesting that they forget it, and advising them not even to muse upon its meaning.It was duly forgotten until late into the twentieth century. The trouble was, no-one remembered or understood what it was really about, and only a few recorded it.

The scene unfolds as follows, arising out of a focus on two characters in the plaza. It begins with a detailed stage direction both in the score and libretto (which is unusual), setting the scene:

The scenic aspect of the first productions of Carmen sprang from impetuses that had burgeoned since the 1830s, fuelled particularly by French and English travellers to Spain who became inquisitive about what they saw as a curiously beautiful, wild and dangerous country. Included among these travellers were many who were in some way involved in the operatic productions. First, there was interest in the legacy of Spain’s buildings, ancient and modern. Second – and this came later – was an interest in its customs, traditions and ceremonies, which were very different from those even in meridional France, and extremely different from those in London.

For these professional travellers – some earning their keep by the pen, the pencil or the brush – Spain began to romanticise itself, particularly by exporting its image in terms of Andalusia, the region furthest from northern Europe geographically and in many other ways, being, as it was, half-European, half-African. In the scenography of the two Parisian set-designers who fashioned Carmen’s first production (1875), and its second staging in 1898 with new sets by Lucien Jusseaume, both features are clear: representation of Andalusian customs is set against the architectural background of Seville and its environs, which opera-goers were invited to travel to in the imagination, aided by realistic sets and, in the outer acts, the ubiquitous tower of the Giralda in the background.

The stage-directions to the opera show a knowledgeable exploitation of the stereotypical customs and costumes documented in jottings and portrayed in images increasingly disseminated from the 1830s onwards. At curtain up, revealing the goings on in an urban plaza, the principal players are the costume-designers and the scenographers, who create an unusually lively tableau vivant of a busy square in Seville, populated with majos and majas engaging in flirtation and banter, and no doubt some petty crime. The music need not be terribly interesting at this point: more important is the audience’s appreciation of this exotic scene: the ‘drôles de gens’ (‘strange people’) neither behaving nor dressing like the French or, for that matter, like the people of Spain’s more northerly provinces. Bizet appropriately provides a few bars of nothing much to go under the customary applause as the opening set is revealed.

On 27 June 1830, as the result of a broken love-affair, Mérimée, aged 27, embarked on a visit to Spain. ‘Les voyages font la jeunesse’, so the French saying goes (‘Travels form your youth’). Certainly, this voyage was to prove crucial: its resonances never left him. As he himself confessed, the final straw was rejection by one Mélanie Double, daughter of a celebrated physician. Mérimée’s attraction to her is documented in letters to Stendhal with whom (among others) he shared intimate details of his sex life. She never married but a chain of letters, now as a confidante, followed. Essentially Dr Double had decided that Mérimée’s financial stability was insufficient for him to marry his daughter. His unluckiness in love, however, turned out to be fortuitous: he admitted that his journey to Spain was ‘the best thing he ever did’. Had he remained in Paris, licking his wounds, we would never have had Bizet’s opera.

Mérimée’s trip turned out to be curative Hispanotherapy, although the Spanish official who had issued his passport had feared that his liberal allegiances, atheism and anti-clericalism might have made his entry into Spain somewhat risky. Just in case, his close friend Stendhal had proposed to Mérimée a Plan B, which was to go to Italy instead. In the end his entry into Spain went smoothly and he soon found himself in the south. Although he visited Spain on several subsequent occasions he was never to visit Andalusia again. This is an important point, since by the time he wrote Carmen he would be drawing upon details of this visit of over ten years before, or else reminding himself of it through reading some of the already-mentioned travelogues. This he certainly did.

Largely by happenstance, he met plenty of like-minded liberals in Spain, not least the Montijo family, to whose children he became known by the pet-name Don Prospero. He had met the brother of the Count of Teba by chance when travelling in a shared diligence, and through this became friends with this famous family, particularly the Count’s wife, Mme de Montijo, with whom he corresponded intimately and frequently, in an entirely Platonic way. Her daughters sat on his lap and loved him. Little was he to know that one of them, Eugénie, would later become Empress of France by marrying Napoleon’s cousin and subsequently settle in England – in Chislehurst, to be precise.

Returning to a reading of Mérimée’s Lettres d’Espagne, less for details that infiltrated both the novella and the libretto, and more for an overall impression of the Spain he visited in the 1830s, we realise that an over-arching sense of the text’s cultures of violence emanates from all four acts of the opera. The extraordinary number of weapons mentioned in the novella, and those retained in the opera, were clearly deliberately inserted to maintain the threat of violence throughout. There is also a sense of the unjust distribution of wealth in Andalusia, which was seen as justifying banditry, smuggling and wrecking. Above all it is clear that this constant background of imminent violence, along with its unpredictability, excited northern travellers to Spain rather than repelling them, and that Mérimée was no exception. He captured this undercurrent in his Lettres and skilfully injected a similar character into his Carmen of ten years later. Study of the libretto reveals that the librettists made some attempt to preserve these darker sides, even though detractors see the opera as a trivialised transformation of Mérimée’s original. In each of the four acts of the opera there are moments of violence either spontaneous or premeditated. Listen to Bizet’s music in that context and maybe the opera appears a little more sinister.

Imminent Violence in Mérimée’s Visits to Spain

As the Lettres evolve there is a distinct move from observation to imagination.Yet even in the description of the bullfight in the first lettre, Mérimée details his emotions as he approaches his first corrida, moving from a fear that he will not be able to endure the ‘free-flow of blood’ to his confession that after he had witnessed the death of the first bull ‘he never looked back’. Of the different types of tauromachic spectacle, he immediately sided with the fully fledged bullfight – which he terms ‘la tragédie’ – rather than the bull-running, such as was (and is) done in Pamplona and the south of France, where there was only minor danger to the men.

Before travelling to Madrid in November 1845, Mérimée had written to his companion and guide Calderón asking him to reserve him a place ‘surrounded by informed aficionados’ at a corrida (if there were to be one during his stay).