On the eve of the 20th century, Alma Mahler wrote in her diary that ‘only one opera exists in the whole world: my Tristan’. A quarter of a century earlier, Clara Schumann had described Wagner’s opera, premiered 150 years ago, as ‘the most disgusting thing I have ever seen or heard in my life’. She wrote further: ‘To be forced to see and hear such crazy lovemaking the whole evening, in which every feeling of decency is violated and by which not only the public but even musicians seem to be enchanted – that is the saddest thing I have ever experienced in my artistic life…During the second act the two of them sleep and sing; through the whole last act – for fully 40 minutes – Tristan dies. And that they call dramatic!!!’ Brahms, meanwhile, claimed that looking at the score put him in a bad mood for the rest of the day.

Elgar, however, when he first heard the Liebestod in 1883, wrote in his concert programme: ‘I shall never forget this’; when he received a vocal score of the work for his 36th birthday he was fully effusive: ‘This Book contains the Height,’ he wrote inside, ‘the Depth, – the Breadth, – the Sweetness, – the Sorrow, – the Best and the whole for the Best of This world and the Next.’ After attending a performance at Covent Garden in 1933, Benjamin Britten reported in his diary: ‘Dwarfs every other creation save perhaps [Beethoven’s] Ninth. The glorious shape of the whole, the perfect orchestration: sublime idea of it and the gigantic realisation of the idea. He is master of us all.’ Grieg, a lifelong admirer of Wagner, if not an uncritical one, couldn’t contain his mirth when his friend Björnstjene Björnson, by contrast, described the opera in a letter to him as ‘the most enormous depravity I have ever seen or heard, but in its own crazy way it is so overwhelming that one is deadened by it as by a drug’. Björnson continued rather less delicately: ‘Even more immoral…than the plot is this seasick music that destroys all sense of structure in its quest for tonal colour. In the end, one just becomes a glob of slime on an ocean shore, something ejaculated by that masturbating pig in an opiate frenzy!’

Just these few choice quotations suggest that no opera, or even musical work – at least before the 20th century – has inspired such visceral and varied reactions as the ‘sublimely morbid, consuming and magical work’ that, according to Thomas Mann, was Tristan und Isolde. Friedrich Nietzsche famously called it the ‘true opus metaphysicum of all art’, writing elsewhere that ‘I am still looking for a work with as dangerous a fascination, with as terrible and sweet an infinity as Tristan – I look through the arts in vain’. On first seeing a score of the Prelude, the philosopher – in what might serve as a warning for anyone setting out to write about the work – reported: ‘I simply cannot bring myself to remain critically aloof from this music; every nerve in me is a-twitch, and it has been a long time since I had such a lasting sense of ecstasy.’

Breaking the rules

The past 150 years are littered with writers trying to express the fascination, revulsion, or both that Tristan inspires; for a further six years before that, we have people trying to fathom the piece either from just the score, published in 1859, or from hearing the Prelude, first performed in Prague that same year and in Paris in early 1860. Wagner sent Berlioz a copy, but even he was left scratching his head by the ‘strange’ first page: ‘I have yet to discover the least idea of what the author wishes to do,’ he wrote. Even today, Tristan remains a work that can inspire fierce devotion or baffled resistance: it eludes clear definition and explanation and encourages intemperate hyperbole at every turn. Maybe Michael Tanner’s thought-provoking description is one of the best: ‘Along with Bach’s St Matthew Passion,’ he writes, ‘it is one of the two greatest religious works of our culture.’

Or perhaps we can turn to more coolly objective matters, and The New Kobbé’s Opera Book, which states that ‘Tristan is the most influential opera in all musical history’. If Isolde herself is allowed in the opera to zero in on the ‘und’, the wörtlein (or ‘little word’) that both unites her and Tristan and places unwanted distance between then, then might we also read significance into the fact that Kobbé talks not of ‘operatic’ history but of ‘musical’ history? The standard assessment of the opera’s role in this latter regard is well known: it broke down the rules of harmony, emancipated dissonance, unleashed atonality and set the foundations, four decades early, for the musical 20th century. If we are to believe for a moment that, pace Robert Craft, Stravinsky’s The Rite of Spring was the ‘prize bull that inseminated the whole modern movement’, then perhaps we can suggest that Tristan provided the potent cocktail that helped get the bull in the mood. The two works could hardly inhabit more different aesthetic universes, but if analysts have tussled for decades to define The Rite’s ‘Augurs of Spring’ chord (before someone suggested that Stravinsky’s hands had simply alighted by chance on a chord that sounded right at the piano), then that is nothing compared with the fascination with the ‘Tristan chord’ – heard almost immediately in the Prelude after the cellos’ first three yearning, ambiguous notes.

The four notes of the chord have been the subject of endless musicological wrangling, which has attempted to define its significance in the opera itself, as well as how it has gone on to have a life of its own, as signifier of heightened and frustrated desire and tension not only in Wagner’s later operas but in all manner of fin de siècle works, good and bad. The New Grove dictionary does its best at a summary: ‘It can be explained in ordinary functional harmony as an augmented (French) sixth with the G sharp as a long appoggiatura to the A, or…as an added sixth chord in first inversion with chromatic alterations.’ If ever an opera seemed resistant to such analysis, though, it is Tristan, whose world is patently not one of ‘ordinary functional harmony’, as is made clear even in those first three bars of the Prelude, whose Langsam und schmachtend (‘Slow and yearning’) marking is as much a precis of the action as a musical direction.

The key signature, ostensibly A minor, gives little away; the 6/8 pulse is all but undetectable. The instrumentation of the chord sets the tone for the score’s entire orchestration: low, mellow wind instruments blend with the cellos to form a new sonority whose exact make-up is difficult to pick out. Throughout the score Wagner exploits this blending of instruments: a technique, pioneered in the organ-like chords of Lohengrin and which later achieved physical realisation in the sunken pit of Bayreuth (Theodor Adorno criticised it in Marxist terms as part of a concern with the phantasmagorical which sought dishonestly to conceal the means of production). But the chord, and the way the following phrase peters out, set the work’s pattern for creating musical expectations that are never resolved, showing why, for Robert P Morgan, Tristan exemplified the most important tendency in 19th-century tonality, by which ‘key centres came to be defined more by implication than by actual statement.’

But it’s more still than that: harmony becomes psychology, the whole score a glorious extended metaphor for unfulfilled desire, and for the philosophical impossibility of fulfilling desire more generally. So complete is Wagner’s achievement in upsetting our harmonic perspective that the C major chords that brassily intrude at the end of Act 1 sound disturbing and disorientating. And when we finally reach resolution at the close of the opera over 5000 bars and four hours of music later, with what Richard Strauss described as the ‘the most beautifully orchestrated B major chord in the history of music’, the final effect can and should be totally overwhelming.

Difficult beginnings

Strauss was a day shy of his first birthday when his father, Franz, played the horn at the premiere of Tristan und Isolde, at the Munich Court Theatre on June 10, 1865. The staging of the opera, six years after its completion, was enabled by King Ludwig II, who had intervened decisively in Wagner’s life the previous year, offering him apparently endless funds (welcome) allied to advice and well-meaning interventions (less welcome). Wagner’s attempt to get the work performed at his own instigation proved fruitless: it famously went through 77 rehearsals at the Hofoper in Vienna in 1863 before the orchestra declared it unplayable. The premiere of the work itself, delayed by a month much to the delight of the hostile elements in Munich, might be counted a modest success. The title-roles were taken by the husband-and-wife team of Ludwig and Malvina Schnorr von Carolsfeld, the former an artist who Wagner admired perhaps above any other singer he worked with.

That the tenor died a matter of weeks after the event, however, only contributed to the legends that surrounded the new work: not only was it incomprehensible and morally reprehensible and dangerous, its detractors noted, it was also literally dangerous. Its moral dubiousness was, and to a large extent still is, underlined by the biographical circumstances of the work’s composition and earlier history, a series of hardly innocent facts to which no one was afraid to add additional untruths. Wagner’s affair with Mathilde Wesendonck in the 1850s, while he was being hosted in Zurich by Mathilde’s wealthy husband (and now usually believed to have remained unconsummated), had inspired the composition. Wagner’s relationship with Cosima, meanwhile, had begun in earnest in 1863 and produced a daughter, Isolde, born exactly two months before the opera was premiered under the baton of Hans von Bülow. Von Bülow was still Cosima’s husband at the time, and the little girl was given his surname, but no one was fooled.

Wagner was disconsolate, depressed by the uncomprehending reactions of the public, but Tristan had at least been unleashed on the world, although it would wait another nine years for a second staging, in Weimar. After Liszt saw it there in 1875 he wrote to Princess Carolyne zu Sayn-Wittgenstein: ‘One felt overwhelmed, ravished, and enraptured all at the same time – in several places one could only weep!’ He added: ‘After so poignant a work I do not know what will be left for our opera composers to do.’ Which brings us back to the question raised implicitly by Kobbé: how did Tristan change opera? Certainly for composers producing operas there was a difficulty of stepping out of Wagner’s shadow, and that of Tristan, perhaps best exemplified by the fact that even while at work on Turandot well over half a century after that Munich premiere, Puccini felt the pressure. He reportedly picked up his score of Tristan, before swiftly putting it back down, exclaiming: ‘Enough of this music! We’re mandolinists, amateurs: woe to him who gets caught by it! This tremendous music destroys one and renders one incapable of composing any more!’

Tristan on stage

No opera produced in a spirit of direct emulation of Tristan has survived in the repertory, while the ramped-up chromaticism of works by Schreker and Korngold – music which perhaps took up the Tristan mantle more than that of any others – often leaves one longing for Wagner’s economy of means. Eventually Strauss found his way of dealing with it through parody, overt or subtle, in the mockingly Tristan-esque duet of Feuersnot, the ‘perverted Liebestod’ of Salome (as Michael Kennedy described the final scene in his Master Musicians volume on Strauss), or the knowing references in Der Rosenkavalier, such as in Octavian and the Marschallin’s own musing on the phrase ‘du und ich’ (‘you and I’) in their post-coital first scene.

Tristan also changed forever what opera could be expected to do, and how it was to be performed. The expectations on the singers are unprecedented, and still to this day Tristans, in particular, seem rarely to be judged by how good they are. The greatest, it seems, are those whose inevitable shortfall in realising Wagner’s unrealistic demands – one is reminded of Wagner’s promise to his publisher that the work would be easy to stage and economically favourable – is the least. In terms of staging, it certainly became clear at the premiere that King Ludwig’s penchant for pseudo-medieval costumes and decorations – given concrete form in the Disney-esque folly of his castle at Neuschwanstein – was fundamentally ill-suited to a work which, according to the German critic Paul Bekker, put ‘sounds not people’ on the stage. Ever since Alfred Roller started to introduce expressionistic touches to his famous designs for Mahler’s 1903 Vienna Opera staging, productions have moved increasingly into the realm of ‘suggestion’ rather than ‘illusion’, to borrow the distinction by the Swiss stage designer and theorist Adolphe Appia. One could spend a lifetime watching Tristan in the opera house today without seeing so much as a ship or castle on the stage.

Here the libretto plays an important part, too: laconic and ambiguous, it often consists of language that writhes tortuously to express the inexpressible, threatening, in the vast love duet of Act 2, to collapse into something like nonsense. Even Isolde’s Narration in the First Act seems to tell a nebulous, dreamlike story. The basics of the plot – putative adultery between Tristan and Isolde, their discovery by King Marke, Tristan’s injury and death – can seem like little more than a simple scaffold around which Wagner was able to build an expression of the negative Schopenhauerian philosophy under whose spell he composed the work in the 1850s. For once, Wagner’s definition of ‘music drama’ – formulated less in seriousness than in a spirit of mocking jest – as ‘deeds of music made visible’ seems appropriate.

Tristan on record

This might suggest that Tristan would be particularly well suited to recording. And bearing in mind Wagner’s letter to Mathilde, in which he expressed concern that a good stage performance of Tristan would be enough to send anyone mad, Tanner suggests that hearing the opera on record has one major advantage: ‘It enables us to stop and wait until we can cope with Act 3; an advantage the tenor singing Tristan must still be more grateful for.’ Right from the very first major studio recording of the work – Fürtwängler for Walter Legge’s EMI in 1952 – the advantages of the studio have been exploited. Famously, Kirsten Flagstad, some way past her prime as Isolde, would only record the role if her diminished top notes could be bolstered by Elisabeth Schwarzkopf. Three decades later, Carlos Kleiber’s DG recording allowed us to hear Margaret Price as the fiery Irish princess, a role she never came anywhere near performing on stage, though we’ll only ever have an imperfect idea of what Kleiber was aiming for in this recording of the sole Wagner opera he ever conducted. Disagreements during the sessions led to him walking out, leaving the project unfinished. DG’s producer, however, had kept the microphones on during rehearsals, and managed to put together a complete performance, released two years after the conductor had abandoned the recording.

Karl Böhm’s famous Bayreuth set, recorded by DG in an empty Festspielhaus in 1966, rattles through the score at some lick, while Leonard Bernstein’s remarkable Philips recording, taped act by act in concerts in Munich in 1981, can seem like an experiment in pushing the boundaries of interpretative possibility. Karajan, in his EMI recording, seems to view the work as a grand symphony, into which the singers often feel poorly integrated, while Antonio Pappano’s 2005 set, also for EMI, is perhaps best enjoyed as an early record of Nina Stemme’s world-leading Isolde (also captured on Marek Janowski’s live-in-concert set for Pentatone). No paid-up Wagnerian would want to be without many of these, or the studio recordings by Solti, Goodall and Barenboim. But it’s surely significant that when Mike Ashman surveyed the work’s discography in these pages (9/06), it was a live recording that came top of the 65 then extant versions: Knappertsbusch’s 1950 set, on Orfeo, from the Bavarian State Opera.

Perhaps this leaves us with one conclusion to draw about Tristan, a work so fascinatingly woven through with contradictions: despite its outlandish demands, it still belongs in the opera house, the very institution it changed forever.

This article originally appeared in the June 2015 issue of Gramophone magazine. To explore our various subscription options, visit: gramophone.co.uk/subscribe