Translators and the performance of language

In the two decades after its premiere, Pygmalion was staged at least twenty times on the Italian peninsula – but exclusively in its original French. During these years, it accumulated about seventeen editions and no fewer than five Italian translations, and yet the text as performed was resistant to the usual domestication. In German cities, by contrast, the work was given in that language.Footnote ¹³ Eventually, Rousseau – or perhaps an editor confronting a half-written sentence – suggested that Pygmalion was designed to accommodate the very un-musicality of its host tongue: opera suited the sonorous Italian language, was even the result of it; French, however, had no fixed prosody and therefore could not generate melodious song.Footnote ¹⁴ But even before this official exegesis became known, there was little room for the work within the categories of Italian opera. For this Pygmalion is a chimera, with parts from different animals – and unlike the dramma per musica, which already boasted a certain hybridity, its unmatched parts were arranged diachronically, in jarring alternation. When Pygmalion made its Italian debut in Naples in 1772, Ferdinando Galiani dismissed the work as ‘half prose, half music, the monster offspring of Rousseau’s imagination’ (despite the efforts of censors, Rousseau was well known by Italians even when he was not loved).

Like many monsters, this Pygmalion was imagined to have origins within the ancient past: an amorphous pre-history in which elements had not settled into their familiar places. Rousseau had been one of many to suggest that language, music and dance shared ancestors in an ancient system of gestures and onomatopoeic cries.Footnote ¹⁵ In ancient Greece these had been fused into pleasing aesthetic forms, such as sung poetry and musical drama.Footnote ¹⁶ Modern historians have perhaps been too eager to dismiss the claim of his musical collaborator, Horace Coignet, that in Pygmalion ‘M. Rousseau wished … to give an idea of Greek Melopoeia, [the] ancient theatrical declamation’; I believe it is quite possible that Rousseau had some kind of revivalist impulse in mind, albeit a complicated one.Footnote ¹⁷ But while a Hellenic musical theatre – such as opera was also imagined to be– could re-emerge in Renaissance Florence without bearing the signs of its age, it could not do so in France nearly two centuries later. After all, Rousseau’s contemporaries were interested in the ruined-ness of ruins: the way that stonecorners eroded, colours faded, statue limbs dropped off and the earth below seemed to be swallowing – ever so slowly – all monuments of human achievement.Footnote ¹⁸

The music of this Pygmalion was made to show its pretended age much as did Piranesian ruins and the plastic works that archaeologists uncovered in their digs – and even (if Rousseau was to be believed) the French language itself. Coignet’s ritornellos replicate the vanishing edges of French words, their silent endings and e-muets, with broken melodies and incomplete cadences. Even the final ritornello (Ex. 1) fails to supply closure. It begins as a gigue for muted strings and horns. But the tendency towards discrete phrases and repetition gradually disintegrates and the harmonies become ever more strange, ending with a half-cadence in E minor – a key that is implied but never heard. As the lovers embrace, the modal shift and chromatic descending bass evoke the chill winds of time and death.

Ex. 1

No. 26, Allegro con sordini, from Horace Coignet and Rousseau, Pygmalion (Lyon, 1770), ed. Waeber; my reduction.

In keeping with Coignet’s assertion of a classical pedigree, Pygmalion’s earliest reception is rife with archaeological fantasies. For instance, the libretto for a 1772 production in Vienna depicted three putti playing with masks and lyres, which they have excavated from a ditch. A shovel lies in the bottom left-hand corner of the image (Fig. 1). What is interesting here is not the putti – the publisher, Josef Kurzböck, used them in most of its designs – but rather the fact that they are shown in an archaeological attitude, uncovering their theatrical toys from below ground. The tablet held aloft by the figure on the right augurs a tabula design within: Kurzböck’s libretto organised Pygmalion’s script into three parallel tracks, separated by vertical lines (Fig. 2).Footnote ¹⁹ On the left are descriptions of the musical interludes, not Coignet’s but rather a new version by Franz Asplmayr. These musical descriptions were probably written for this production, without Rousseau’s involvement, but they circulated more widely than Coignet’s. The middle column of the libretto records the length of the interludes (again not traceable to Rousseau), while the column on the right alternates between passages of gesture, prescribed in italics, and the monologue in regular typeface. This table form made the text seem unfamiliar by supplementing and classifying its uses: though Rousseau’s register within Pygmalion had been, in the end, bourgeois and mundane, the Viennese design recalled transcriptions of ancient or distant languages.Footnote ²⁰ Yet as Kurzböck’s putti showed, even very old objects could occasion new practice once they are found: to that end the publisher printed Italian and German translations alongside the original French, and the work was performed in German. Taken as a whole, the libretto asked the spectator to observe a variety of forms of translation, between languages as well as between words and other media – audiences were invited to read (as it were) the gesture that translates language, and the music that translates gesture.

Fig. 1

Title design, Pygmalion, Kurzböck edition (Vienna, 1772).

Fig. 2

Three-column libretto, Pygmalion, Kurzböck edition (Vienna, 1772).

The first Pigmalione published on Italian soil made a similar invitation to its readers.Footnote ²¹ This was the work of Francesco Saverio de’ Rogati, a Neapolitan classicist who had attended the Naples premiere in 1772.Footnote ²² Though Galiani suggested that the only amusement on offer that evening was Galatea’s ‘very interesting body’, de’ Rogati had admired both the declamation and Asplmayr’s expressive interludes.Footnote ²³ His 1773 Pigmalione appeared in the same dual-language format – original on the left page, Italian translation on the right – that had graced his editions of Sapphic and Anacreontic poems.Footnote ²⁴ Here again, the work is presented as debris. The title of de’ Rogati’s Pigmalione appears on a stone tablet with degraded edges, half buried (Fig. 3); in the background stands three columns of a temple ruin.

Fig. 3

Title design, Pigmalione, trans. Francesco Saverio de’ Rogati (Naples, 1773).

Finicky and expensive, the dual-language format was common in Italian publication at this time only when the age or distance of the original seemed to require it. Here, though, de’ Rogati did not offer a translation between remote languages, but rather an adaptation between similar languages with different esprits or geni.Footnote ²⁵ Rousseau’s fractured prose is fashioned into the shape of an operatic libretto, with versi sciolti for recitative and versi lirici for arias. De’ Rogati promised his readers that he ‘endeavoured to supply the public with an exact poetic translation … never moving far from the original’, granting only those concessions that were required to render prose into poetry.Footnote ²⁶ These in themselves were substantial. For instance, the French prose:

becomes, in Italian poetry:

As this passage makes clear, de’ Rogati did not preserve the original disjointed style. Ellipses are gone and fragments are woven into coherent thoughts; for instance, ‘vanity’ and ‘human weakness’ – unmoored in Rousseau’s French – are accounted for here as ‘appropriate for such a beautiful work’. However, and despite inversions in the destination language, most of Rousseau’s words are present in something like their original order (here alone we may note vanité to vanità, puis to posso, ouvrage to opera, enivre to ebro, and amour propre to proprio amore): they settle as kernels of the original within a more prolix unfolding.

De’ Rogati made his purpose clear in the preface: Pigmalione was to be the prototype for a new kind of Italian opera. He alluded briefly to the faults of opera seria, but cited his friend Saverio Mattei, who described them at greater length.Footnote ²⁷ Their critiques were very much like those of Francesco Algarotti, Cesare Baretti and Calzabigi from the 1750s and 1760s: the conventions of opera seria were, after all, still robust in the mid-1770s despite the appearance of all Gluck’s Italian reform operas in the preceding decade.Footnote ²⁸ To begin, in the list of conventional operatic faults there was the endless series of solo numbers: Metastasian libretti usually featured only two or three duets or larger ensembles amongst more than a dozen exit arias. Then there was the much lamented da capo, which called for the first verse to be repeated after the second even when it was nonsensical to do so. The inertia of the musical form was aggravated by the habit of most singers of remaining stationary as they sang. What was more, the standard two verses of an aria were too short to accommodate the appropriate amount of music, encouraging composers to write long orchestral introductions and repeat words and phrases ad nauseum. But most pernicious of all was the coloratura: the rapid scales and arpeggios that showed off voices at the expense of prosody. Such text setting within arias created a stark contrast with the intervening passages of recitative, which had a more natural pace of declamation but were musically uninteresting. Indeed, accounts suggest that Italian audiences paid little attention to these portions of an opera, and composers often assigned the task of setting simple recitatives to a pupil.Footnote ²⁹

This is not to say that the Neapolitan critics cared for melodrama per se. Indeed, de’ Rogati shared Galiani’s low opinion of it: ‘One cannot imagine a greater lack of verisimilitude than accompanying or preparing with an instrumental ritornello the declamation of an actor who does not sing but speaks.’Footnote ³⁰ And yet the actor who played the hero – Louis Bursay, who had also starred in the 1772 Viennese production – seemed to capture the audience’s attention in a way that Italian opera only rarely did: ‘the precision, the vivacity, the passion of the very able actor, the appropriateness of the few musical notes of Mr. Aspelmayer [sic], distributed in time and sensibly, awakened in the audience a greater delight than the recherché, long, and ever dull music of today’s Italian theatres’.Footnote ³¹ De’ Rogati asked composers who set his Pigmalione to preserve something of Bursay’s manner of speaking, so that ‘the song itself is not much different from French declamation’ (‘il canto, presso a poco, non sia più lungo della declamazione Francese’).

In other words, de’ Rogati wished Italian song to sound more like French speech, and it could be modelled in the first instance on his own translation of a French text. His Pigmalione is not quite one of those ‘monsters of language’ (mostri linguistici) abhorred by purists: the style of the Italian still remained a safe distance from the French; in this matter at least de’ Rogati was a conservative. Rousseau had admired the prosody of Italian: within the realm of opera, whatever the faults of its practice, the Italian language was as imperialist as Napoleon would later be. But he could also have mentioned, as many of his compatriots did, that centuries had passed since that language had thrived outside the opera house.Footnote ³² Calls for a reform of Italian via the reception of French thought sounded with increasing urgency as the century progressed. At mid-century, a moderate such as Giuseppe Baretti promised in his translation of Corneille to avoid ugly francesismi: ‘words and phrases of the humble French language’ in the midst of ‘words and phrases of noble Tuscan’.Footnote ³³ But in the 1760s and 1770s even some Italians declared their language to be archaic, tolerant of unnatural syntax and thus apt for poetry only.Footnote ³⁴

These literary debates are well known; what is not known is that they played out on stages as well, within trans-linguistic theatrical performances, and the adaptation of melodrama in particular. De’ Rogati’s libretto was never set to music; as we will see, though, his principle of an Italian melody that sounded like speech – even foreign speech – had gained traction by the time Pimmalione finally arrived on the operatic stage in 1790. In the interim, Italian actors followed the model of Bursay by performing Pygmalion in French. It was, of course, not uncommon that French dramas should be given in the original language in Italy. It was strange, though, that Rousseau’s scène lyrique was performed in French by Italian actors, with libretti that featured eminently performable translations and proclaimed the mutual foreignness of text and actor. At the centre of this practice was the actor and playwright Tommaso Grandi, a lightning rod for French influences within Italian theatre during the 1770s and 1780s. Touring with the work between 1773 and 1781, Grandi had occasion to commission at least four editions of Pygmalion and two different translations. Each of these libretti advertised the production in the same way, emphasising the foreignness of Pygmalion’s words within the performers’ mouths: ‘Pimmalione, by Mr. J. J. Rousseau … presented in the French tongue in the theatre of this city, by Tommaso Grandi and Teresa Monti, Italian actors.’Footnote ³⁵ Others followed his lead. In 1778, Neapolitans could observe Pigmalione, ‘represented in the French language, in prose, by Giacomo Ceolini, Italian actor’.Footnote ³⁶ Meanwhile, Pygmalion was being performed by ‘Bonifazio Welenfeldt and Annette Paganini, Italian actors’ during the middle of the decade. Indeed, Welenfeldt, who hailed from Parma, seems to have had a talent for performing across this language divide: within improvisatory theatre he was known for the character of the ‘Italianised Frenchman’.Footnote ³⁷ The reference within such libretti to unspecified civic theatres suggests – despite the vaunted philosophical provenance of the original – a measure of itinerancy reminiscent of the Commedia dell’Arte; and indeed, the careers of Monti, Grandi and Welenfeldt had their origins there.Footnote ³⁸

Such performances required new types of supplementary printed text. Grandi was almost certainly responsible for a 1773 performance of Pygmalion at the Teatro San Samuele, done in French and probably with Asplmayr’s music. The title page boasts a ‘New Translation, to which is added the Original by the Author’; both languages are presented in columnar form.Footnote ³⁹ Grandi himself might have been the translator, or he might have asked his some-time collaborator Elisabetta Caminer Turra.Footnote ⁴⁰ The libretto included a few prefatory remarks on the difficulty of the task, and a humble admission that this Italian Pimmalione contained ‘only a trace of what was brilliant in the original’.Footnote ⁴¹ Modern observers might be inclined to disagree on both counts, as the drama has merely been given a faintly Italian patina: the opening line ‘il n’ya point là d’ame ni de vie’ becomes ‘La non v’ha nè anima, nè vita’; ‘ce n’est que de la pierre’ becomes ‘essa non è pura che pietra’; and the translation continues word for word to the end, with few surprises.

The idea of a French original was so essential that one had to be invented for the first new melodrama created in Italy. Orphée, which had its premiere in Venice in 1774, had a French prose libretto by ‘L.B., pastore arcade’ and music by an unnamed composer.Footnote ⁴² It was performed by its author/translator, who explained that this new scène lyrique was made in Pygmalion’s model: ‘you desired another scene of this genre; I present to you this Orfeo … Between the immortal work of the celebrated Rousseau and this feeble imitation there is no similarity other than that of genre.’Footnote ⁴³ Yet there is an abundance of resemblances between the two works, and it is not clear which are fundamental to the ‘genre’: they stage neighbouring passages within the Metamorphoses; both contain artist-heroes and revived heroines; both are spoken in French, with parallel Italian for reading only.

What is more, Orfeo prescribed gestures and actions as well as spoken text, with ‘music [that was] analogous to the pantomimic expression of the actors’.Footnote ⁴⁴ This principle of a music-gestural analogue came to the earliest melodrama from pantomime dance, both rhetorically and in practice.Footnote ⁴⁵ Within pantomime, bodily gesture and musical gesture were construed as two parts of a composite sign system that could be more expressive and accurate than verbal language – if, that is, the two streams were calibrated to be perfectly simultaneous.Footnote ⁴⁶ Pantomime dance and melodrama also shared performers and musicians. Franz Asplmayr, whose Viennese Pygmalion score was most often used in Italy, also wrote music for the ‘ballets d’action’ of Georges Noverre.Footnote ⁴⁷ Many of Asplmayr’s interludes suggest an obvious correspondence between on-stage event and musical event, and there is little doubt that the two media were designed to coincide exactly. (This is particularly true of the shorter interludes.) Elsewhere, though, his score for Pygmalion suggests a more complex relation between motion and music than is usually attributed to early melodrama: particularly when vestigial dance tropes create a fleeting rhythmic momentum.

For instance, Asplmayr’s second interlude (reproduced here as Ex. 2) must analogise a series of actions and gestures: the sculptor ‘tosses away his tools with disdain, becomes restless, walks, stops, reluctantly turns his gaze toward the bottom of his atelier where the pavilion hides a statue, turns away, and falls into a deep reverie’.Footnote ⁴⁸ The throwing of the tools probably coincided with the first chord, which is made more prominent than the subsequent chords by its low bass note. Here the left-most column of the Viennese libretto, devoted to musical descriptions, is helpful: it tells us that ‘the music expresses with rapidity the first few of these movements’. The footsteps of the actor’s promenade thus probably coincided with the crotchets, while the syncopated treble notes conveyed agitation and perhaps, as in French pantomime dance, affective upper-body movement.Footnote ⁴⁹ But the syncopations soon give way to a six-bar sequence, coloured rather prettily with dissonant passing tones; the schematic chords and arpeggios are replaced by a music that is driven by its own tension and resolution, as the crux of the hero’s inner conflict is suggested by a glance towards the statue Galatea. Perhaps paradoxically (at least for the modern listener), as Pygmalion ‘falls into deep reverie’ the orchestra shudders into silence; this is, after all, the music of audible motion, not inwardness.

Ex. 2

No. 2, Allegro, from Franz Asplmayr, Pygmalion (Vienna, 1772).

These music-gestural interludes provided further opportunity to observe translation, quite literally in action, and Rousseau’s Italian disciples made the most of Pygmalion’s ties to the language of pantomime. In July 1775, for instance, Grandi and Monti’s performance of Pygmalion was followed by a complete rendition of the same work entirely in pantomime dance.Footnote ⁵⁰ This translation was the work of choreographer Antoine Pitrot. Grandi/Monti’s and Pitrot’s double Pygmalion functioned as a single entr’acte for Domenico Poggi and Antonio Salieri’s comic opera La locandiera. It would seem that no libretto or music survives from the dance half of this entr’acte. However, as Pitrot was a disciple of Noverre, there would have been three main components to his adaptation: excising Rousseau’s spoken words in favour of yet more gestures, choreographing the appropriate footwork and commissioning a composer to supply a continuous, danceable score. The evening’s entertainment – which would surely test the patience of any audience – supplied yet another form of analogue: sign systems were placed side by side, and audiences invited to judge the distance between them. This time, of course, the translation was ‘read’ not over a page margin but through a comparison of consecutive performances.

Similarly, in Naples in 1785, when Domenico LeFevre choreographed another pantomime Pygmalion, he designated it a scène lyrique (following Rousseau) despite the absence of words altogether. Summarising his ballet within the libretto, he marked the animation passage with a footnote describing the care he took to ‘translate’ Galatea’s first words of self-identification (the celebrated moi and n’est plus moi) directly into the appropriate gestures: ‘Here I imitated J.J. Rousseau, as I did whenever Galatea repeats this word positively or negatively. This expression is so simple, so natural, and depicts Galatea’s situation with such truth, that I believe that the effect it creates to be infallible, whether rendered in words or in gestures.’Footnote ⁵¹ A model translator in these details, LeFevre allowed himself some liberty with a few others; for instance, he supplied a vigorous thunderbolt to animate the statue and appended a long divertissement for a cast of dozens.

But surely the strangest of these entertainments came from Count Alessandro Pepoli, an aristocrat who performed the title role in a French-language production of Pygmalion at the Accademia de’ Rinnovati in 1788.Footnote ⁵² His ‘favola odecoreutica’, Ati e Cibele, featured a speaking heroine and a hero who can only mime.Footnote ⁵³ The music was contributed by another aristocratic amateur, Giambattista Cimador, who will become a major character in the third act of this present study. The year 1792 saw the premiere of Pepoli’s Piramo e Tisbe, in which a singing hero is divided from a speaking heroine.Footnote ⁵⁴ Both of these works replicated Pygmalion’s chimeric nature with similar (and similarly jarring) alternations of theatrical means; in both, characters who seem to have wandered in from different genres must attempt to understand one another across an idiom divide. Pepoli’s interest in translation went beyond the stage. Seeking to import French thought into Italian culture, he devoted his Tipografia Pepoliana largely to providing Italian readers with an annotated Biblioteca teatrale della nazione francese: twenty-seven volumes of plays by Voltaire, Molière and even Quinault, translated by Pepoli, Melchiorre Cesarotti and others.Footnote ⁵⁵ Rousseau’s Pygmalion provided a consistent centre to Pepoli’s various activities during the last ten years of his life, 1787 to 1796; he even wrote one of the first original Italian melodramas, Pandora.Footnote ⁵⁶ As we will see, Pepoli – like de’ Rogati – sought finally to revitalise opera itself, by means of a new kind of song modelled on the pitches and rhythms of speech.

Mimophones and phonographs

But what did speech actually sound like, and why did it sound the way it did? These questions were asked frequently during the 1770s and 1780s, in Italy and elsewhere, and Italian opera reformers who were interested in the sounds of utterance could draw on a variety of sources. Rousseau himself forcefully affirmed that words exist first and most meaningfully in sound; a language haemorrhages truth and sentiment when it is written in alphabetic characters.Footnote ⁵⁷ In trying to imagine how words were spoken (or sung) more than two hundred years ago, historians are now little better off than Rousseau had been in confronting the ancients. The sounds themselves are long vanished and their notation ever incomplete. If absolute answers are few, some relative ones may still emerge from the comparisons that were made between idioms and languages. For instance, when one of Pygmalion’s Italian translators claimed to have captured ‘only a trace of what is brilliant in the original’, he indulged in a particularly modern fantasy about textual fidelity, especially since these remarks prefaced a word-for-word conversion. Such fantasies were rare within Italian culture of the late eighteenth century – even Melchiorre Cesarotti’s celebrated translations now seem to take too many liberties – but they were unheard of in Italian opera and pantomime, which operated with the most liberal principles of adaptation. These fantasies are rendered even stranger by the proximity of the French and Italian languages. If ‘Tire, ville opulente et superbe’ had significance that was not preserved in ‘Tiro, opulente e superba Città’, then meaning must have been heard in minute gradations of verbal sound.

What is more, the categories of speech and music are slippery within the original melodrama. Pygmalion declines to specify pitch and rhythm for its most exposed medium: the actor’s voice. Did this mean that the actor simply uttered the text, as he might do today in spoken theatre? Not necessarily: the melodramatic voice has been musical in more recent memory, and I would suggest that it was this way from its inception.Footnote ⁵⁸ Even Rousseau (or, again, his editor) suggested in the ‘Observations sur l’Alceste italien de M. Gluck’ that the sculptor’s declamation should have a musical component, but an improvised one: ‘a sensitive and intelligent actor, by bringing together the tone of his voice and the accent of his declamation with what the musical line expresses, mixes these foreign colours’.Footnote ⁵⁹ A listener who was unmusical might not hear the difference between such sonorous declamation and true recitative. The sensitive listener would notice merely a slight, persistent dissonance between the vocal pitches and the orchestral.Footnote ⁶⁰

The difference between speech and song is one of degree rather than kind: speech also has pitch and unfolds in time; it has, in other words, a melody and a rhythm. The ‘speech-likeness’ of the spoken word comes largely from the fact that its pitches are not gradated but continuous, and they are determined spontaneously, usually without regard for the musical logic of motives or patterns.Footnote ⁶¹ These elements varied (as Rousseau and his contemporaries were vividly aware) with the idiom and register of the utterance: poetry was more melodic than prose, as the patterns generated by rhyme and metre created something like a musical structure. Vowels could be open or muffled; some languages had fixed long and short syllables, making their ‘feet’ tidier; some languages’ diacritical marks indicated the raising and lowering of pitch.

The various melodies of language coexist uneasily with the idea of music as a literate practice. Ancient Greek, for instance, had an abundance of musical qualities; this very fact may have discouraged the Greeks from developing more thorough notation, which could have preserved their music for posterity. So, at least, believed Joshua Steele, whose Prosodia rationalis (1775) is among the oddest eighteenth-century treatises on music. Steele maintained that the imperfect notation of the ancients had resulted in centuries of quarrels about the precise sounds of their language.Footnote ⁶² In order to ‘treat the modulation of speech as a genus of music’, he invented a new system to record all aspects of the spoken word.Footnote ⁶³ Steele insisted that his speech notation could be ‘read off as easily as a song tune’. It used time signatures, staves and Italian expressive terms (staccato, piano, forte and so on). However, by dispensing with noteheads, he was able to mark the tiny gradations of tones used in speech. In Figure 4, the lines attached to the bottom of the stems represent pitch variation, while the symbols at the top provide rhythm; the squiggles trace dynamic swells, while the commas signify a piano dynamic. His treatise includes ‘scores’ for the opening of The Iliad, as well as Hamlet’s monologue ‘To be or not to be’, with annotations permitting readers to pronounce it as had the actor David Garrick.Footnote ⁶⁴ Garrick’s influence is evident throughout the treatise: in all of his figures, Steele also provided ample rests for breathing, which he called ‘stops of expression’ because ‘a pause or silence, fitly employed, makes a significant impression on an auditor’.Footnote ⁶⁵ The sustained cello tones in Figure 4 are unique to this example: according to his experiments, the bass drone ‘animates the reader or speaker’, causing him to stay closer to the initial pitch of a syllable – as suggested here in ‘end’ and ‘aim’ – rather than sliding directly away as shown at the syllables of ‘happiness’.Footnote ⁶⁶

Were this system to be widely adopted, Steele boasted, ‘the overwhelmed ruins of a Herculaneum should be retrieved from rubbish, and restored to their former splendours’.Footnote ⁶⁷ Unlike most, he believed that the English were the truest heirs to the Greeks – a conviction that was assisted by his unusual ability to detect perfect metrical feet within his own language. His argument distilled into a single maxim: ‘every sentence in our language, whether prose or verse, has a rhythmus peculiar to itself; that is, in the language of modern musicians, it is either in common time or triple time; videlicet, minuet time, or jigg time, or mixed’. The dictum was designed to supply its own proof in his notation (see Fig. 5). The utterance begins in duple metre and then settles into triple; note the puns at ‘minuet time’ and ‘jigg time’.Footnote ⁶⁸

Steele’s notation could also preserve moments of great eloquence for posterity. No less a figure than Lord Monboddo – author of The Origin and Progress of Language – praised him for having ‘fallen on a way to make Garrick live as long as his Shakespear’.Footnote ⁶⁹ This aspect of the project did not escape notice by another of Steele’s contemporaries. In a footnote within the Life of Johnson, James Boswell wrote that Johnson’s ‘mode of speaking was indeed very impressive; and I wish it could be preserved as music is written, according to the ingenious method of Mr Steele, who has shown how the recitation of Mr Garrick, and other eminent speakers, might be transmitted to posterity in score’.Footnote ⁷⁰ If Steele’s system had been employed in a timely fashion – Boswell noted – it could have been used to discredit those actors (like a certain Mr Henderson) who imitated only imperfectly what Boswell called Johnson’s ‘deliberate and strong utterance’, and others dubbed his bow-wow way.Footnote ⁷¹

As Steele’s ‘General precept’ demonstrated, heightened or aphoristic speech had its own nested medium of icons; that is, its pitches and rhythms contained representations that are analogous rather than arbitrary. This ideal formed the basis of one of the most influential works of Italian music theory to appear during the second half of the eighteenth century: Antonio Eximeno’s Dell’origine e delle regole della musica (1774). Its author was not a musician, but rather an ex-Jesuit mathematician who had been displaced to Italy following the expulsion of the Jesuits from Spain. Like many fraters, Eximeno was living out his exile in Rome, where he could attend the services at St Peter’s. As the papal choir sang the Mass, he covertly spoke it along with them ‘with the same energy that [he] would have used to move the congregation to devotion’. One Pentecost, Eximeno noticed that his recitation of the Veni Creator Spiritus was identical in rhythm and melody to that version by Niccolò Jommelli that the choir sang. Thus came a realisation pivotal to Eximeno (though hardly novel in music theory): ‘Music is but a prosody to give language grace and expression.’ Therefore, he argued, music should be derived neither mathematically nor contrapuntally, but rather from the pitch variation and rhythm of common speech (il commun parlare).Footnote ⁷² The difference between speech and song lay in the pronunciation of vowels, and particularly the most sonorous: O. This letter also had a pleasingly pictorial notation: ‘O is a drawing of the position of the mouth required to pronounce it, i.e., concave and open’ (p. 150). In the cave of the mouth’s O, breath vibrated and echoed rather than exiting immediately as with the other letters; it thus seemed to be sung even within the context of speech.

Eximeno thought that languages usually aged much like stones. Most people in his day had collapsed mouth-caves, which rendered their voices unmelodious and muffled. However, Eximeno believed that he had found in his adopted homeland a magical site of semi-preservation: something like a living fossil in sound. Italian (he proclaimed) was second only to ancient Greek in being naturally melodious, because Italians retained open caverns in their mouths. Cicero suggested that Roman women pronounced Latin the most sweetly, and Eximeno was happy to announce that the practice persisted, ‘the influence of the climate on their language more durable than the stones of Campidoglio’ (p. 409). Thus while in Italy one could hear music in the streets: ‘without exaggeration’, he stated, ‘it can be said that the Italian speaks as he sings. Even the most colloquial speech is never without its distinctive cadence’ (p. 409) – the diatribe of an angry Roman woman was an operatic aria.

The song within Italian speech supplemented verbal meanings with mimetic sound. Eximeno repeated the truism that the first humans communicated by vocal object-imitation: ‘in the first instance (if other circumstances permitted), instinct caused the inflections of the voice to express the qualities of the object’ (p. 135). Phonic mimesis, like gesture, declined with the development of modern languages – unless, that is, the decline could be staved off by means of an exquisite natural sensitivity. Because ‘his soul is extremely receptive to the most delicate sensory impressions’, the Italian retained mimetic systems alongside the verbal; for instance, the word ferro resounds sharply, while aria causes breath to rush softly between the lips (p. 135). Each vowel had a distinct character. Furthermore, the Italian ‘spoke’ in movement, reinforcing verbal meanings with face, hands and arms.Footnote ⁷³ These movements formed the basis for Italian pantomime dance, but could also be observed in the most colloquial utterance. For instance, Eximeno reported having ‘the pleasure of observing, in a cafe in Rome, the loveliest sight in the world: a natural mute, sitting in a circle and conducting a long conversation, he and everyone else speaking only with gestures’ (p. 411).

Eximeno’s treatise is primarily contemplative, but others directed his conclusions towards musical practice. Esteban de’ Arteaga, the most prominent opera critic of the day, cited Eximeno when he suggested a new basis for operatic reform: ‘a dramatic system … built on the exact relation of the movements of the soul with the accents of the word and of language, and [the exact relation] of those accents with musical melody’.Footnote ⁷⁴ As standard practice had changed little in the intervening years, Arteaga’s complaints about the opera of 1785 were very similar to de’ Rogati’s and Mattei’s in the early 1770s: he decried the harsh divisions between recitative and aria, the ritornellos, fragmentation and repetition of poetry, and excessive coloratura. According to Arteaga, song should be assembled from the composer’s memory of passionate utterances, theatrical or real, while preserving those ‘inflections and accents’ originally given to each word as pronounced. The fragments of remembered sounds should be woven together according to the rules of melody. In this sense, its principle of imitation was no different from that of sculpture, for instance, which fashioned a single human out of parts from several.Footnote ⁷⁵

Yet musical representation could even be superior to that of the plastic arts – Arteaga claimed – provided it used the voice in the proper way: ‘song is the most complete and most interesting imitation that the fine arts can achieve … Painting and sculpture can only imitate, as one says, the skin of man; song can penetrate right into his soul, awakening its movement and depicting its most intimate changes.’ Such pre-eminence arose from the twin nature of song’s imitation, preserving the pitch-iconic within the verbal: ‘imitating directly the tones of human language, the very elements of song that name the represented object also function as a means of representation’. In other words, the operatic voice could be both plastic and animated – as Arteaga illustrated with the imagery of a familiar fable:

But some languages (like some materials) were inhospitable to the crafting of these seductive images. Italian was suitable for song because of ‘the correspondence between the sounds of its words with the nature of the object they express’. Operatic prosody should be deduced from a continual observation of ‘the inflections made by man in his ordinary voice’; on this matter, Arteaga referred his readers to Eximeno’s treatise. From common speech also comes ‘the origin and force of [song’s] imitation’. Yet (for Arteaga as for Eximeno) Italian ‘familiar discourse’ promised an even grander union of the arts: native speakers reinforced verbal meanings with the face, hands and the rest of the body.Footnote ⁷⁷

Arteaga claimed to derive ‘the nature of opera’ from ‘the hidden wellsprings of philosophy’, and he blamed the genre’s current decadence on a lack of philosophy amongst composers.Footnote ⁷⁸ In this at least he found an ally in Alessandro Pepoli, whose open letter to Calzabigi in 1788 announced the need for ‘a music controlled by philosophy’ (una musica diretta dalla filosofia), and lamented the same set of offences within practice.Footnote ⁷⁹ But Pepoli decried Arteaga’s conservative taste in poetry, asserting that no new opera could be born of the old Metastasian librettos. Instead, composers required new dramas in flexible forms, which dispensed with the old alternations of poetic and speech-like registers in favour of a continuous unfolding. Composers must be guided by the ‘force, truth, and necessity of imitative music’ (viii); the song itself should be ‘controlled by a philosophical economy’ (x), while requiring the actor to engage also in ‘the science of mime’, la scienza mimica (vi). After their experiments with translation and hybridity in the late 1780s, Pepoli and his circle at the Rinnovati turned to the problem of transforming Pygmalion into an Italian performance text, finally (nearly two decades after de’ Rogati’s first call for melodrama-inspired reform) turning its French speech into something like Italian song. As melodrama was finally adapted into a sung idiom, the rhetoric of Greek revival flourished around Italian opera one final time, reborn as a phoenix from the ashes.

Pimmalione

Rousseau’s Pygmalion arrived on the operatic stage in Venice, at the Teatro San Samuele, on the evening of 26 January 1790.Footnote ⁸⁰ It featured two of the era’s foremost stars: the title role was given to the opera-seria tenor Matteo Babini, while Galatea was a ballerina, Carolina Pitrot. (Pitrot’s father Antoine had choreographed the pantomime for Florence’s double Pygmalion in 1775.) The libretto included a letter to the audience signed by Babini and Pitrot, outlining the terms of the adaptation: this Pimmalione had been tailored for Venetians and ‘only the principal idea of its celebrated author’ had been preserved.Footnote ⁸¹ This preface does not signal that the scène lyrique has become an ordinary opera, as some have suggested.Footnote ⁸² Rather, it echoes conspicuously the rhetoric of failed adaptation that prefaced the dual-language Venetian libretto of 1773, used by Tommaso Grandi, on which this later adaptation was based. The 1790 libretto was actually quite faithful. In both cases, the translators’ fidelity occasioned a series of departures from standard practice: while the 1773 libretto contained a mostro linguistico of Gallic deformities, the later adaptors fashioned a hybrid opera from a host of foreign elements (not to mention performers from different idioms). They coined a new designation, scena drammatica, declining again to stake a claim to comic or serious register – a notable omission even within this period of genre multiplicity.

The 1790 libretto was the work of Simeon Antonio Sografi, a friend of Pepoli and student of Cesarotti himself. While it was based on the 1773 edition, which had also been created for the San Samuele, Sografi followed the form of the original Pygmalion and preserved a good deal of its language. Most of the prose becomes blank verse, denoting recitative, while the stage directions are translated wholesale. For instance, in the beginning of the scene, gestures are rendered verbatim, and almost every distinctive feature of the French prose has been delicately modified into Italian poetry:

Sografi, Pimmalione (Venice, 1790)

Rousseau, Pygmalion (Vienna edition, 1772)

Pimmalione seduto, ed appoggiato sopra il gomito si và atteggiando a guise d’Uomo inquieto, e melanconico. Si alza risoluto, prende i suoi Strumenti, e tratto tratto con lo Scalpello ritocca gli Abbozzi. Si allontana da essi, e li garda con afflizione, ed avvilimento.

Pygmalion assis & accoudé rêve dans l’attitude d’un homme inquiet & triste: puis se levant tout à coup, il prend sur sa table les outils de son art, va donner, par intervalles, quelques coups de ciseau sur quelqu’une de ses ébauches, se recule et regarde d’un air mécontent & découragé

A che spirto, nè vita

Il n’ya point là d’ame ni de vie…ce n’est que de la Pierre…je ne ferai rien de tout cela –

Più darvi non poss’io.

Dove sei genio mio!

O mon genie, où es-tu?…

Che mai sei divenuto

mon talent

Misero mio talento!

qu’es-tu devenu?…

In te tutto è già spento

tout mon feu s’est éteint…

Quel foco animator, ch’opre immortali

mon imagination s’est glacée…le marbre sort froid de mes mains…Pygmalion, tu ne sais plus des Dieux…tu n’es plus qu’un vulgaire artiste –

Facea sortir un dì … Itene al suolo

Voi strumenti non più della mia Gloria,

Ma del mio disonor. Lascia tu pure

Avvilito Scalpello

Vils instrumens qui n’ètes plus ceux de ma

Questa mano volgar; non sei più quello.

gloire, allez…ne deshonorez plus mes mains.

Getta con dispregio i suoi Strumenti.

Il jette avec dédain ses outils.

The translation approaches word for word in many places: Rousseau’s line ‘O mon genie, où es-tu?’ is rendered by Sografi as ‘Dove sei genio mio!’; ‘mon talent qu’es-tu devenu?’ similarly receives only a delicate Italian patina, becoming ‘Che mai sei divenuto/Misero mio talento!’; the line ‘Vils instrumens qui n’ètes plus ceux de ma gloire’ is rendered as ‘Voi strumenti non più della mia Gloria’; and so on.

Pygmalion is inhospitable to the most common sort of lyricism, the exit aria: no character in this drama ever exits the stage. Instead, Sografi distributed five cavatinas amongst the recitative. The first, ‘Voi che intorno a me vi state’, is the longest, with two stanzas of four ottonari and a concluding stanza of two, each preceded by mime:

These lines are necessarily further from Rousseau’s than was the blank verse. This aria encompasses roughly from the direction ‘Il s’assied, & contemple’ to the line, addressed to Galatea, ‘je serai consolé’. But for the first two stage directions, the hero’s basic emotional trajectory and a few prominent terms (ardeur–ardore, impétueusement–impeto, tourment–tormento, console–consolar), the two passages have few similarities. Yet Sografi maintained dramatic momentum even within the lyric register: there is (in other words) no stopped time or ‘aria time’ that is distinct from ‘recitative time’.Footnote ⁸⁴ The poem spans three states of mind: first Pygmalion’s sculptures calm him, then he becomes agitated, and finally he admits that he longs only for Galatea’s statue. Thus this progress through Pygmalion’s thought resists the return of a first verse or musical A-section: a principle followed by all of Sografi’s cavatinas, with one important exception. He concluded the work with a duet: while Rousseau’s Galatea did little more than utter her famous moi, this statue coquettishly requests a lesson in love.

The task of setting Sografi’s Pimmalione fell to Giambattista Cimador, who had composed the music for Pepoli’s Ati e Cibele the previous year. Indeed, Cimador’s main credential was a familiarity with the dramatic principles of the Rinnovati; he had no particular facility with the materials of opera, as voice-leading errors and awkward modulations attest. In other words, his composition is a clear instance of what Pepoli called ‘music directed by philosophy’, and here philosophy prescribed continuity above all. Where the melodramatic Pygmalions were simply collections of musical fragments, riven by speech and silence, Cimador’s Pimmalione is preternaturally free of fissures: the entire work is one single movement, without discrete numbers, without coloratura or recitativo semplice, and, for the most part, without strong points of cadence. The terseness of this music is evident from the opening bars, as is its tendency to favour brief transitions over points of closure. Like the libretto, the score labels particular gestural segments from ‘Scena I’ to ‘Scena XXI’.

The vocal line contributes to this sense of an overarching smoothness. Cimador set all the versi sciolti as accompanied recitative: sometimes the voice sounds amidst sustained string chords, while elsewhere the orchestra supplies a more active texture of harmonic support and interjection. This alone is sufficient to merit the work a place in music history, as perhaps the first Italian opera to forego simple recitative altogether since the reform of Gluck and Calzabigi. Except for the continuous presence of orchestral sound, Cimador’s recitative resembles most Italian recitative of that century, complete with the old Handelian cadence formulas – a point worth underlining lest the notion of ‘numberless opera’ might prompt imaginings of more recent musical idioms.

But this recitativo accompagnato permitted Cimador to interweave the song with instrumental passages for the mime. Take, for instance, the final words of the opening monologue, followed by the first gestural interlude, and then the hero’s resumption of speech (Ex. 3). The last words of ‘Scena I’ – ‘non sei più quello’ – describe a typical recitative cadence, with a falling leap of a fourth in the voice followed by a dominant chord. The resolution doubles as the opening of the next section: a passage of mime. Here, the hero ‘throws down his tools with disdain; walks agitatedly, stops, and turns as if by force towards the bottom of the stage, immediately withdraws his gaze, and falls into deep meditation’.Footnote ⁸⁵ It is, in other words, the stage direction whose corresponding musical number by Asplmayr we examined as Ex. 2.

Ex. 3

End of Scena I, Scena II, from Antonio Sografi and Giambattista Cimador, Pimmalione (Venice, 1790).

Here again, the miming devices are easily spotted: the G minor chord on bar 1 of ‘Scena II’ coincides with the throwing of the tools; a bass line that moves on every crotchet illustrates the ‘walking’, while the faster-moving upper parts supply ‘agitation’. In the seventh bar of ‘Scena II’, syncopations become a questioning arc of crotchets, and the walking bass is replaced by the first of several B-flat semibreves; this must indicate the moment when Pygmalion ceases his steps and turns toward the veiled statue of Galatea. The last five bars provide this short mime with a final dominant-seventh chord built on B-flat, briefly suggesting a key of E-flat minor that ultimately never arrives. The hero ‘falls into meditation’ with a slowly descending melody; his inner unrest is made audible with a half-step oscillation in the bass, as the accompanied recitative begins once more.

In other words, Cimador used the usual techniques of gestural mirroring. Indeed, he may have modelled these portions on the parallel moments within Asplmayr’s score, which Pepoli had almost certainly used for his 1788 performance of Pygmalion in French at the Rinnovati. Certainly, the similarities between Exx. 2 and 3 are striking: the Allegro tempo and C minor harmony; the ‘walking’ crotchets in the bass offset by syncopations in the upper parts; a second texture with melodic motion on the beat and one bass note per bar once the walking stops; falling melodic motion to indicate ‘falling into meditation’; and finally the prolonged dominant, inflected by an augmented sixth chord, in the concluding bars.

Most of Cimador’s Pimmalione sounds much like the music in Ex. 3, alternating between accompanied recitative and mime. Some of gestural segments are given discrete sections of music enclosed by double bars, while others coincide with briefer interludes, even just cadences. What, though, of the cavatinas? How could they be made to sound like speech? Lyricism poses the most obvious threat to the project of an Italian opera built on French prose; after all, Rousseau’s hero had no recourse to messe di voce or roulades, and nor did his speech alternate between irregular patterns and regular ones. To avoid jarring shifts of register Cimador set each lyric verse in a continuous iteration, with no fragmentation or repetition. (There is one exception: the last two lines of verse are sometimes repeated with new music in order to supply a fleeting sense of closure, as had been a common French practice since Lully’s early operas.) What is more, this Pimmalione declaims his poetry syllabically, moving between adjacent pitches within a narrow ambit: the rôle spans no more than a seventh from beginning to end. The musical metre unfolds in lockstep with the poetry – verbal accents invariably land on the first beat of each bar – and thus the lyric form generates the musical form. The opening melody of the first cavatina is reproduced here in Ex. 4; in these bars, the poetry unfolds at a pace that is almost mathematically regular. Note values are confined within a narrow rhythmic range: most are crotchets or quavers, with the occasional semiquaver or dotted crotchet.

Ex. 4

Beginning of Scena III, from Sografi and Cimador, Pimmalione.

For each of the remaining two stanzas Cimador supplied a new musical character, complete with new key and time signatures, all the while following the same principles. The result is a through-composed form of three very small linked sections, with the small bridges between verses mirroring the actor’s gestures. Like the speech melodies within Steele’s Prosodia rationalis, Cimador’s music is essentially a-motivic: because this music is, in principle, determined at every moment by the text, whether gestural or ‘spoken’, every event is isolated and unrepeated. The listener is hard pressed to find a single reprise, and even sequences are rare. If Rousseau’s hero is unable to complete sentences, Cimador’s has a profound suspicion of points of arrival: in the broadest of terms, sentimental ellipses translate to open cadences. ‘Voi che intorno a me’ is exemplary in this regard: it contains only one point of musical arrival, marking the last syllable of the last word, and even this repose is undermined right away by yet more accompanied recitative.

Although unlikely to feature in a History of Musical Beauties, Sografi and Cimador’s Pimmalione was the first attempt to introduce the text and techniques of Rousseauian melodrama into something like an operatic genre. What is more, it was both influential and extraordinarily popular, remaining in the repertory for nearly thirty years. Matteo Babini toured with Pimmalione throughout the 1790s, performing it ‘to astonishing effect’ in Venice, Bologna, London, Vienna, Florence, Genoa, Paris and even further abroad.Footnote ⁸⁶ Several more composers set Pimmalione in Cimador’s model. There were new versions by Francesco Sirotti (Livorno/Milan, 1792), Francesco Gnecco (Genova, 1794), Bonifazio Asioli (Turin, 1796) and Francesco Moro Lin (Venice, 1797) in the 1790s alone. These scenes use the same stylistic palette: syllabic declamation, recitativo accompagnato throughout, cavatinas, discrete miming segments, minimal points of musical arrival and an ever-present impulse towards through-composition.

But the influence of melodrama extended beyond settings of Sografi’s libretto. At first, Pimmalione’s musical techniques remained bound to the themes of co-dependent media and monologic indecision. Its imprint is evident in the mime-interspersed operatic soliloquy of the 1790s, in which singers performed prescribed, audible gestures as they confronted statues, pictures, or letters. A comic instance can be found within La statua per puntiglio (The Stubborn Statue, Venice, 1792), which parodied Rousseau’s drama within a stock commedia plot.Footnote ⁸⁷ The Dottore figure, Tolomeo, is an antiquarian, swindled by his adversaries into buying a variety of false animated figures. Near the end of Act II he awakens in front of a statue that is made to speak (Pygmalion-like, but again falsely) later in the opera. There follows an extended segment of mime, granting Tolomeo his own animation scene: ‘at the sound of a soft instrumental interlude, [Tolomeo] slowly awakes, and expresses with attitudes appropriate to his circumstance the surprise that he feels at seeing the objects before him’.Footnote ⁸⁸ In the Camilla by Giuseppe Carpani and Ferdinando Paer, the object that provokes recourse to gesture is not a statue but a letter: ‘The duke orders his writing desk to be brought. He begins to write. He tears up what he has written. He takes the portrait out of the desk and gazes at it. He kisses it. He holds it to his heart. He sighs. He slams shut the writing desk and rushes off.’Footnote ⁸⁹ In Sografi’s La morte di Cleopatra (Venice, 1794), Cleopatra finds a small vessel containing the fatal asp, and subsequently enacts parallel deaths in mime and speech.Footnote ⁹⁰ In these moments, the object seems to confer its mode of being (though perhaps briefly) onto the protagonist, making him into a cypher. This reflexivity of substance is evident in the most canonic of animation scenes; after all, did not Don Giovanni have a ‘cuore di pietra’, and was not Rousseau’s sculptor himself animated by desire?

Of course, such Pygmalionesque monologues within opera had no extended moments of orchestral silence, nor did they feature the dissonant slides of vocal pitch that indexed historical age for the original melodrama. In other words, these Venetian fusions of singing voice, acting, and music did not seek to represent ruins. Instead they were promoted by a rhetoric of classical rebirth: perhaps the last resurgence of an Arcadian impulse in opera. In 1792 the city gained a new opera house – La Fenice (the Phoenix) – whose very name bears the traces of this near-forgotten impulse. La Fenice’s inaugural opera seria, Paisiello’s I giuochi d’Agrigento, had a libretto by none other than Pepoli himself. One of the most distinctive features of this movement was to cast its singers as bards, who made music spontaneously as they declaimed. In Platonic terms they were purveyors not just of mimesis but also of diegesis. A portrait of the tenor Matteo Babini – who (we recall) created the title role in the 1790 Pimmalione, and who also appeared in the central role in this work – is inscribed, in Greek script, ‘a complete bard, the first in an age’.Footnote ⁹¹ If Babini were hardly the creator-musician that this praise implied, his admirers could at least assert that composers were obliged to write for him in a style befitting his talents. For instance, his biographer insisted that Babini had asked Cimador and Sografi to write Pimmalione in ‘precisely’ the way that they did.Footnote ⁹² The other serious roles written for him lend some weight to this assertion. For instance, Brutus in La morte di Cesare (Gaetano Sertor and Antonio Bianchi, 1789/97) and Orazio in Gli Orazi e i Curiazi (Sografi and Marco Antonio Portogallo, 1798) sing plain melodies within shorter forms.Footnote ⁹³ The opening of Brutus’s second-act cavatina ‘Voi che il cor a me vedete’ is strikingly similar to Cimador’s ‘Voi che intorno a me’, beginning with a lyrical woodwind solo and proceeding from there with a syllabic rendition of the first few lines. Yet Bianchi did not sustain such extreme declamation for long: a second time through the text contains four opportunities for cadenzas, and concludes with a flourish of coloratura for voice and English horn a due.

This bardic imagery implies that the actors composed their own roles. Few did, of course, but Pepoli was again an exception: in 1793, Venetian audiences were given the opportunity to admire yet another setting of Sografi’s Pimmalione, with orchestral music by Giuseppe Rossi but with vocal melodies written and performed by Count Alessandro Pepoli himself. These melodies do not survive, but we can be confident that they were ‘generated by a philosophical economy’ – a foundational principle for Pepoli’s reformed opera.Footnote ⁹⁴ In a dedicatory letter to his 1793 libretto Pietro il Grande, Sografi summarised Pepoli’s ideal of ‘musical declamation’, claiming that he had written the title role for Pepoli in order to ‘demonstrate the use that could be made of his example by many of our so-called star singers, both through song and through dramatic declamation, too often overlooked’.Footnote ⁹⁵

By this time Sografi was Venice’s premiere librettist, sought after in the other Italian capitals.Footnote ⁹⁶ His libretti after Pimmalione had a similar disregard for the formal procedures that had regulated lyric numbers and set pieces. With a willing composer, the results could be striking – sometimes to the displeasure of audiences. For instance, his 1793 collaboration with Niccolò Piccinni, Ercole al Termodonte, alternated parlante lyricism, gestural music and long portions of accompanied recitative. The score made a convenient marriage between Italianised melodrama and Italianised tragédie lyrique. It was pulled from the stage of the San Carlo after four performances. Sografi had more success with the dramas he wrote for Venice, which could usually rely on Babini, and whose composers were more selective in their stylistic transgressions. For example, in Portogallo’s setting of Gli Orazi e i Curiazi, Orazio’s ‘Se alla patria ognor donai’ provided a mostly syllabic text setting, but the vocal melodies are showier than in Pimmalione, their rhythms martial rather than prosodic. ‘Se alla patria’ alternates between accompanied recitative, choral interjections (a requisite of larger dramas in the Pepolian model), and brief cavatina for Babini. Portogallo’s most striking innovation here was to set versi lirici as a new kind of accompanied recitative, marked ‘ad libitum’ and ‘come a recitativo’ with the orchestra playing ‘colla parte’.Footnote ⁹⁷ This did not imply vocal embellishment, as ‘ad libitum’ often did (for instance) when it was marked at a cadence. Rather – and this is a striking intervention indeed – Portogallo was telling his principal singers to improvise their own melodies for lyric verse, and asking that these should be made to sound like recitative.Footnote ⁹⁸

But melodramatic ideals of gesture and declamation could prove still more disruptive to operatic conventions, particularly when librettists liberally interspersed their lyric stanzas with gestural motions requiring musical analogues. A prime example may be found within L’amor coniugale (1805), a one-act ‘sentimental farce’ with a libretto by Gaetano Rossi and music by Simon Mayr.Footnote ⁹⁹ Like Pimmalione, it adapted a French play, in this case Léonore, ou l’Amour conjugal (1798) by Jean-Nicolas Bouilly; here again, sentimental prose is rendered mostly as syllabic Italian song. Mayr had given some attention to the Italian offshoots of melodrama: he was in Venice when Cimador’s version of Pimmalione had its premiere, and later assembled a nice collection of Pimmalione libretti and scores.Footnote ¹⁰⁰ One of L’amor coniugale’s chief attractions is the monologue ‘Sì, ne profitterò … Rendi il consorte amato’, in which the heroine Zeliska resolves to liberate her husband from prison, hears the bell toll four (as counted on her fingers), and then rushes offstage to make the rescue. Rossi supplied a few lines for recitative, followed by a long stretch of settenari. These twenty-two lines of versi lirici are interrupted more often by ellipses and prescribed actions than the preceding recitative – and unlike in Sografi’s Pimmalione such interruptions do not respect regular patterns. Gaetano Rossi was never the most elegant of librettists, but these are impossible to scan, and some are closer to ten or eleven syllables than seven (as for instance in the penultimate line, ‘Vola il tempo … Oh Cielo, oh amore!’). Indeed, when the opera was revived in Venice in 1808, the engravers mistook most of the passage for recitative.Footnote ¹⁰¹

This monologue received some of Mayr’s most Cimadorian music.Footnote ¹⁰² The opening recitative is all accompagnato, with one brief orchestral interlude noting the instant at which the heroine ‘turns to the heavens with great passion’ (volgendosi al Cielo con maggior fervore). At the beginning of the recitative a three-note motive repeats periodically, and then a two-note one, but even this technique soon dissolves. The lyric section begins with a restrained cavatina, marked ‘Larghetto cantabile’, metrically regular but through-composed. No portion of melody is repeated, and there is only one complete cadence. Most notable for our purposes is the final movement, a perpetually failing cabaletta: its texture shifts continuously, it has no ritornello, no opening melody or recurring motive, little repetition and a text setting that is syllabic until the final lines, often sounding more like an accompanied recitative than an aria. The density of gestural portions poses the greatest strain on form, as the orchestra is made to act during these moments as a descriptive medium rather than a self-sufficient one.Footnote ¹⁰³ Even as the rhythmic momentum grows, the orchestra continues to shift from one gesture to the next, eschewing those devices – like repetition or cadential closure – that might signal a design larger than the local requirements of text. Its materials are those clichéd devices of musical mimophony: rising scales, sighs, palpitations, sudden shifts between loud and soft. By the end, Zeliska’s hysterical melodic leaps and decisive phrases strain against a motiveless music and near metre-less text.

The dream of an opera written by singer-actors did not outlast the eighteenth century. However, through Babini’s influence these Venetian experiments created a new kind of tenor hero, whose song amplified an (imaginary) utterance within spoken theatre, and for whom acting included a substantial amount of mime. Giacomo David also performed Cimador’s scene during the 1790s and later. The role of the sculptor was taken up by Giacomo’s son Giovanni David, and by soprano Marianna Sessi in London, Paris, Berlin and Prague during the 1810s. If Babini’s most famous student, Gioacchino Rossini, decisively rejected Italian declamatory song in the 1810s and 1820s, many bel canto singers continued to cultivate the earlier style, using the earlier music, to demonstrate their acting talents. Giovanni David sang Cimador’s scene in a concert for his own benefit at Naples’s Teatro nel Fondo in 1818; also featuring on the programme was the third act of Rossini’s Otello, in which David had created the role of Rodrigo two years earlier. Such practices kept declamatory opera in the active repertoire after audiences had lost their taste for it. The Rossini tenor Nicola Tacchinardi created his own scena lirica – so he labelled it – by excerpting a monologue from Mayr’s opera I misterj Eleusini.Footnote ¹⁰⁴ The version by Asioli was revived as late as 1831, its statue none other than Maria Malibran.

Sografi’s Pimmalione libretto was set to music a final time in 1816, by Gaetano Donizetti at the prompting of Mayr, his teacher. This scene follows the model of the earlier versions: accompanied recitative and mimed passages alternating with short lyric portions, and with a declamatory text setting. Donizetti discovered Rossini’s music soon after completing Il Pigmalione, and never again composed in this style, nor on another ancient topic. Yet he had an enduring interest in the musical depiction of onstage movement. Here it will suffice to name a few in one of his mature operas, Maria Stuarda (1835).Footnote ¹⁰⁵ The Act I duet between Leicester and Talbot begins with an orchestral section representing the departure of Elisabetta, Leicester’s approach to Talbot, and a handshake between the two men. Donizetti depicted the promenade much as Asplmayr had sixty years earlier, with ‘walking’ crotchets and syncopations; the passage is given a certain coherence with a semiquaver motive, Mayr’s technique (pp. 67–8). Similarly, in the Act I finale Maria kneels before Elisabetta, and a repeated three-note figure descends from the upper strings to the violas and cellos (p. 288). In Act II, fittingly, the most conspicuous miming music sounds when the queen hesitates over Maria’s execution order: as she takes up her pen to sign, the orchestra descends regally down the E-flat major triad; as she halts, the tutti chord ceases and the strings sound a series of quiet tremolos down the chromatic scale; finally, as she leaps to her feet, the winds re-join the strings for a surprise forte (pp. 381–2). These techniques were resilient in their plainness – and perhaps for this reason Donizetti confined them to recitative and as connective tissue between scenes. They do not, in other words, decimate his cabalettas.

It is in the music of Vincenzo Bellini that we may hear the speech-like song advocated by de’ Rogati and Pepoli in its purest form. This lineage is most evident in his early operas, and was identified immediately by their startled spectators, who called the style ‘philosophical’.Footnote ¹⁰⁶ His operas for La Scala bear a number of strong family resemblances to the first essays in melodramatic opera. Through large portions of Il pirata (1827) and La straniera (1829), Bellini employed vocal lines that were so austere, so resolutely speech-like, that one contemporary critic wondered whether they should be called canto declamato or declamazione cantata: spoken song or sung speech.Footnote ¹⁰⁷ An emphasis on ‘the rhythms of the spoken word’ above all resulted in brief lyric portions that eschewed ornamentation and repetition, taking their melodic forms directly from the verse, precisely as had Cimador’s. Perhaps more striking, though, are the fluid musical forms that play host to this song. Bellinian ‘numbers’, like Gualtiero’s aria in Act I of Il pirata, feature a continuous texture, devoid of extended ritornelli but defined instead by a constant interchange between accompanied recitative, arioso and orchestral gesture; a single number might encompass dozens of tiny shifts of character, tempo and key, and a few larger ones. (John Rosselli called this music ‘startlingly Wagnerian’.Footnote ¹⁰⁸) We may now understand that when Bellini’s contemporaries called his style ‘philosophical’ opera, they were casting it as the heir to that Pepolian operatic revival, academy-born, which proclaimed the historical importance of Calzabigi and Gluck but found more tangible inspiration in Pygmalion.

Orpheus diffused

There is one final musical consequence of these projects that is worthy of notice: another thread that can be traced between the first all-sung Pimmalione and the annals of nineteenth-century music. We may catch a first glimpse of it near the end of Cimador’s scene, when the otherwise steadfast principles of adaptation are briefly suspended. In Sografi’s libretto, just before the animation, the sculptor extends his hands upward and prays – a moment that has no parallel within Rousseau’s text:

With his plea a portal seems to open. As Pygmalion stills, the strings suddenly go silent in medias res and there sounds a ‘soave Armonia’: solo oboe, bassoon and harp (Ex. 5 reproduces the final bars of this interlude, and the entry of the voice). This music is heard not only by the audience, but also by the sculptor himself:

Ex. 5

Excerpt from Scena XVI, from Sografi and Cimador, Pimmalione.

Fig. 4

‘Oh, happiness!’ Joshua Steele, Prosodia rationalis (London, 1775), 38.

Fig. 5

‘General precept and example’, Steele, Prosodia rationalis, 28.

Pygmalion’s song is suddenly no longer isolated musically from the orchestra. Sometimes the voice takes its motives from the solo oboe and bassoon, sometimes it is repeated by them, and occasionally the parts move in unison or thirds with one or another. Each line is incomplete, but together they form a single, diffused melody.

These moments bring a third character just within audible range of Pimmalione’s diegesis. It is not Venus (Rousseau’s fable and its adaptations had little use for her); rather, it is the original bard, Orpheus himself. The coaxing harp arpeggios overheard by Cimador’s hero seem a direct reference to those played by Gluck’s Orpheus as he calmed the furies at the beginning of Act II (or the parallel scene within Bertoni’s adaptation of the same libretto). Orpheus was a submerged presence within Book Ten of the Metamorphoses: he narrates the fable of Pygmalion alongside the partner tale of his own Eurydice.

Who was the Orpheus made audible here? Several participants had claim to the laurels. One was the storyteller, Rousseau; he had, after all, confessed to having taken up music in order to become a ‘modern Orpheus’.Footnote ¹¹¹ Another was Cimador, who, after all, wrote the music. Still another (as we have seen) was Babini himself, whose proponents cast him as the ancient Greek bard returned in the flesh. This proliferation of Orpheus figures brings to light the thorny question of agency within these projects. Goethe himself considered this problem in Wilhelm Meisters Lehrjahre. In the passage in question, the hero is travelling with the actor Laertes and his troupe when they encounter a mysterious old harpist, who entertains them with a song. Wilhelm is provoked into an ecstatic meditation on the representative powers of this kind of music as compared to live theatre. He suggests that there was ‘more live presence in [the harpist’s] singing than in our stiff stage personages’, and such musical narratives might even replace actors. Laertes objects that acting is easier when music is involved. Normally the actor must control his own declamation, timing and gestures, and resist being disturbed by the other performers; in a music-theatrical work, though, he ‘becomes another person’, giving over his agency to the music and allowing it to ‘enliven’ and ‘control’ him – it determines his movements, the ‘manner of delivery’ and the expression.Footnote ¹¹²

In a recent landmark study by Mary Ann Smart, the passage provides a point of departure for a progressive loosening of the bonds between music and gesture, which occurred over the course of the nineteenth century.Footnote ¹¹³ So let us conclude this study by reconsidering it as a point of arrival: an articulation of the concerns of those late eighteenth-century projects that subsumed speech and movement within the domain of music. Goethe knew Rousseau’s Pygmalion and had written a small piece, Proserpina, in emulation: during the 1770s itinerant actors in the German territories – much like the fictional Laertes – had performed the original melodrama much as performers such as Tommaso Grandi had on the Italian peninsula. When Laertes claimed to surrender his agency within musical theatre, he almost certainly spoke of melodrama and its offshoots. In practice, musical miming figures such as Asplmayr’s ‘walking’ crotchets were indeed agents of external control, setting the footfalls into rhythm like a flat-footed vestigial dance, the limbs translating gestures scripted for them. The actor moved when the music moved; he stopped when it stopped. He was set into motion from outside.

Thus, while the passage ostensibly contrasts music-theatrical mimesis and sung diegesis, ultimately it considers both to follow a single model: in both, a narrative agency controls subordinate images or characters. In melodrama and its offshoots, the actors were manipulated by a narrative agency that is both omniscient and legible: Rousseau’s Italian admirers circulated Pygmalion’s libretto as a grid-like apparatus in order to bear witness to this very ideal. Conversely, the ancient or revived musical voice contained its own ‘presences’ or actors, its icons and onomatopoeia; thus Arteaga could boast of song’s animated statues, the ‘most perfect imitation there is’. If melodramatic projects destabilise the agency of the performer, they invite a parallel reconsideration of the composer’s role. As we have seen, such musical icons were often antagonistic to autonomous musical forms: how, for instance, does one admire Simon Mayr’s skill in turning out a gestural passage? It requires a leap of imagination at the least – or, better yet, a performance. Melodramatic music becomes mundane in the absence of visual counterparts, and this very mundaneness has militated against an appreciation of its historical significance, and the myriad threads that run between Rousseau’s project and the nineteenth-century operatic stage. After all, melodrama is ever inert on the page, like the torn fragment of an enchanted cloak that becomes, as it falls to the ground, a piece of the dreamer’s pyjamas.