Interview, David Trippett on Liszt's Sardanapalo

Today sees the release of the world premiere recording of Franz Liszt’s projected opera Sardanapalo, which the composer embarked upon at around the same time as his son-in-law Wagner was putting the finishing touches to Lohengrin but sidelined after the near-completion of just one act. Following extensive reconstruction and orchestration of the manuscript, the existing material was premiered in Weimar last year and recorded for posterity; in the run-up to the recording’s release, I spoke to the musicologist and driving-force behind the project David Trippett about the process of bringing this music to life for the first time, and the mysteries surrounding the genesis and eventual abandonment of the work…

It's incredible that this piece hasn’t attracted attention before now – what put you onto the trail, and why do you think it’s been neglected for so long?

I think it’s one of those cases where the music has simply sat and waited patiently for its discovery! The manuscript was catalogued fairly early on in the twentieth century, but was dismissed by some very important Lisztians as being fragmentary and more or less illegible, and these proved to be influential comments. Liszt isn’t known as an opera composer at all, so perhaps that also contributed to its neglect.

As a student in Leipzig I came across a brief reference to the score in an article, and as I wasn’t far away from Weimar I jumped on the train to have a look. What was immediately clear was that Liszt had written out a lot more than had been suggested: in fact all the vocal lines were complete and continuous. There appeared to be gaps in the score, but on close inspection it became clear that these weren’t actually missing sections: this wasn’t someone writing sketches in the way that (for instance) Beethoven did, with themes, counter-themes and germs of ideas for later development. Liszt worked by improvising at the piano and in his head, so it’s fair to say that when he began notating works on this scale they had already been through a lot of preparation in his mind, and by the time he reached the writing stage he already had a continuous, complete conception. (That’s true of seven out of the nine sketchbooks in Weimar, of which this is one). I felt that, at the very least, something could be rescued for a song or aria, but it was only when I had some research time a few years ago that I thought: ‘Actually, he'd worked this out very carefully – we just haven’t been reading it in the right way’.

How much work did you have to do to prepare a performing edition from the manuscript?

I think the main reason that people hadn’t taken this forward in the past is that there are places where the accompaniment simply drops out, so it’s easy to assume that Liszt just hadn’t thought it through. But if you look above that, the vocal lines are not only continuous but they’re often complete with ornamentation, phrasing, and the text above them. It does seem a little odd that a musician as prodigiously productive and intelligent as Liszt would write out a really text-crafted, harmonically thought-through vocal line and not bother thinking about what the accompaniment would be. But these gaps only occur where there’s a recurring accompanimental pattern; in Italian opera you often have rather formulaic accompanimental patterns, and I think he simply felt: ‘This is so obvious that I don’t actually need to write it out – any competent assistant could do this’. (There's a whole critical discourse about this aspect of Italian opera: the Germans thought it was very uncreative, and Liszt made precisely this criticism in the late 1830s when he was travelling in Milan and went to La Scala). So that was the main editorial leap of faith that I needed to make, recognising that where’s there’s a gap it wasn’t due to indecision: he just felt that it wasn’t worth his time.

The libretto is a completely different kettle of fish, largely because there’s no separate text. The underlay in the music manuscript is the only source, and we had the added problem that Liszt didn’t actually speak much Italian: he loved Italian poetry and culture, and he quoted Dante and Petrarch in his letters, but he wasn’t attuned to the finer details of the language. Another difficulty occurs where the text underlay simply disappears. The easy cases to solve are those where he indicates that there’s a repetition of text (which is quite normal in this genre), but there are also instances where that’s just not the case – so we had to reverse-engineer things and use the rhyming and metrical scheme to supply our best guess as to which word he didn’t write out. It’s possible that the libretto might turn out in someone’s attic at some point, of course, and I am in touch with colleagues at Sotheby's and other auction-houses to see if there’s anything on the radar…

Does the manuscript provide any clues as to orchestration?

We know from correspondence that Liszt was going to give this score to his assistant in Weimar, Joachim Raff, to orchestrate, as was his pattern with symphonic works. His method was to write out a hybrid vocal/short score, sometimes just on two staves but expanding in places to five, where he separates out vocal line, woodwind, strings, brass, percussion. The biggest clue is that throughout he provides specific cues (sometimes he’ll specify violins with mutes, in the trio he indicates timpani, and in other places it’s more general cues like ‘woodwind’), so he was clearly already thinking in orchestral colours. But this information is not consistent, and sometimes entirely lacking, so what I did beyond those cues was to take the various operatic and symphonic scores that were on Liszt’s desk in the early 1850s and look for similar musical material, using that to guide the orchestration. These included Lohengrin and Tannhäuser of course, Benvenuto Cellini, Fidelio, and earlier works by Donizetti and Bellini which he had absorbed as a travelling virtuoso, and studied for his own piano paraphrases and transcriptions.

There are also five or six places where he very specifically identifies how accompanimental patterns should be divided between instruments – for instance that a figure should be split between first and second violins. These were all instructions for Raff, and it’s incredibly fortuitous that he prepared the manuscript in this way, because it supplies latter-day scholars with a good amount of primary information, allowing someone in 2018 to produce something that’s as close as possible to what he had in mind in 1852. He would have made changes to whatever Raff produced, and of course we can’t possibly know how far that would have gone, but there’s certainly a good amount of evidence there.

You mentioned Italian opera, and there seems to be a certain amount of ambiguity in his attitude towards the genre – how far in this score do you think he moves away from the language of Donizetti, Bellini et al?

One review of the premiere said ‘It’s amazing how fluent Liszt is in his Verdi’: the opening chorus is very Italianate, and Mirra’s huge scene has a clear cavatina, slow movement, cabaletta etc.. But I think Liszt was trying to modernise the Italian genre, to monumentalise it as a kind of modern drama. In 1854 he published twelve essays about the history of opera, in which he analyses works like Meyerbeer’s Robert le Diable in some detail, and talks about the legacy of Metastasio, Piave and Scribe. He felt that what was unforgiveable in some Italian opera was a kind of brutalising of Shakespeare or Hugo, disrespecting the integrity of the drama - he really gets quite upset about this, and it seems that Sardanapalo was his attempt to translate one of his great heroes, Byron, into a present-tense musical setting. He felt that the future direction for opera would rely on more declamatory melody, greater stage realism – so we find some broken vocal lines, with rests inserted in between otherwise legato statements that really force the singers to be a bit more declamatory in their utterance. But at the same time you have these ravishingly beautiful bel canto lines, and I think the sheer extent to which Liszt became embedded in this aesthetic really surprised Kirill [Karabits] and the singers: it’s a real cornucopia in that you have some almost traditional Italian sections sitting alongside music which leans much more towards the likes of Wagner. You might call is a kind of through-composed bel canto. Liszt was of course the inventor of symphonic poetry, the premise of which is bringing together great literature with music, and this score shows how he attempted exactly that in a different medium; one could even say that the whole enterprise of symphonic poetry—telling a story or depicting a scene in the present tense through music—was actually kindled within the aesthetic of opera. You only have to think of the Dante Symphony, which premiered in 1857 and was actually conceived as a multimedia, staged show.

Speaking of Wagner, does their correspondence indicate whether they discussed the project?

They were certainly in touch about it. Wagner writes to Liszt in 1849 to goad him into producing an opera: he says ‘[Seeing your incidental music for Goethe’s Faust is like seeing] the claw by which I recognise the lion – but now I call out to you, Show us the complete lion!’. I don’t think Liszt had told him that he was working on this project, and he writes back a fortnight later saying ‘[Actually,] my opera will be completely finished in the summer!’. Wagner was initially very encouraging (remember he was quite dependent upon Liszt for political support), then writes back after the opera failed to materialise and recommends that Liszt stop trying to compose an Italian opera and instead compose a German opera for Weimar. Wagner was really trying to pressure Liszt into setting his own cast-off libretto for Wieland der Schmied, but faced with Wagner’s rather knotty German alliteration Liszt may have felt that the Italian opera was the easier option! He also greatly admired what Wagner was doing with German opera, and didn’t want to become involved in that at all: ‘I must abide by my resolution never to compose a German opera’. I don’t think Wagner ever heard this music, but it’s intriguing to consider what he would have made of it.

Do you think that dynamic between the two played a role in Liszt’s ultimate abandonment of the project?

That theory’s certainly been put forward, but personally I don’t think Wagner played a role in Liszt’s decision: Liszt was a extremely famous, wealthy, an important figure; he was the senior partner in that relationship, so I don’t think that he was kow-towed by Wagner’s slightly manipulative sentiments. Wagner’s Opera and Drama included a virulent diatribe against Italian opera (where he states that ‘with Rossini opera died’ and so on); I don’t think Liszt had read that whilst he was composing Sardanapalo, but even after he did, his plans for the opera continued: we know that at the end of 1851 the orchestration was still on the horizon.

I think there’s a very practical reason why Liszt didn’t complete the work, namely that he never received a revised libretto for the second and third acts. He almost certainly never knew the name of the librettist; he’d initially approached the French poet Félicien Mallefille, and even paid him some money up front, but Mallefille missed a couple of deadlines, and Liszt became very frustrated and fired him. (He tells a friend: ‘Our Shakespeare will not or cannot complete the task. To hell, then, with Mallefille's Sardanapalo!’, and that sort of anger is unusual for Liszt, who was generally a diplomatic and generous person). The exiled Italian princess Cristina Belgiojoso then procured a libretto from someone who—it seems—was imprisoned for agitating for Italian independence against the Hapsburgs, and drafted the libretto in jail. This person clearly knew how to write opera: there’s is a slightly self-consciously archaic flavour to some of the language, but in terms of conventions, it’s a perfectly decent libretto. We just don’t know who he was: I wondered for a time if it could have been Agostino Ruffini, but I don’t think that’s right. And not being able to rely on a named librettist might have contributed to Liszt’s lack of confidence about the libretto and its drama; eventually the trail goes cold, and he became so busy with other projects that this one just stalled, fatally.

Can we assume that Liszt wasn’t writing with specific singers in mind in the way that Bellini and Rossini often did?

It’s possible, but he doesn’t specify individuals. We know that he originally wanted to premiere this at La Scala, then he thought about Vienna, then Paris, then London. Given that all of these houses had different house singers, I think it’s unlikely that he had particular people in mind. And though Liszt regularly attended the opera and rubbed shoulders with Rossini etc., not working regularly in opera meant that he didn't have the experience and contacts of a Donizetti or a Rossini - but he absolutely wanted the best singers in Europe, and the work needs a very good soprano, so perhaps he would have approached someone like Henriette Sontag or Giuditta Pasta. He occasionally provides up to three different versions of ornaments in the vocal lines, and offers an alternative—busier—line at the end of the cabaletta: Joyce [El-Khoury] and I agreed that there’s an almost instrumental virtuosity in the rapid triplet leaps that he expects, which are the kind of things that a well-schooled pianist would manage comfortably, but it’s a different matter for the soprano! Hearing the score with voices was a step which Liszt never took, and if he had done, it’s possible he may have revised and adapted.

Actually taking that step of introducing singers into the mix on his behalf was incredibly exciting: we rehearsed and performed just a stone’s throw from his blue composing room in the Altenburg in Weimar, and it was an incredible experience to hear the score with singers for the first time. Even the orchestral musicians felt that excitement: one of the violinists said to me in rehearsal: ‘You know, I never particularly liked Liszt, but after this I’m fully converted: I think I’ll give him another chance!’