Like most very great artists, Stravinsky was by trade a synthesiser of experience. Works that seem to display the sublime detachment of pure art (whatever that may be) turn out to have complex origins in the social and intellectual life around him, origins which could hardly be guessed from the music but which, once known or suspected, can help us come to terms with complexities or even obscurities in the work itself.

There is nothing new about this; the nineteenth century offers several well documented parallels (Schumann, Wagner, Mahler). All the same, the case of Stravinsky is in some respects unique. The Romantics thought of themselves as outcasts, but Stravinsky really was an outcast. Stranded in Switzerland when war broke out in 1914, he was finally cut off from his Russian homeland by the Bolshevik revolution of 1917, and thereafter he made only one trip to Russia, in 1962. He moved from Switzerland to France, then later (in 1939) to the USA, and materially speaking he settled wherever he lived: he was no spiritual outcast, like Mahler, nor consciously part of a diaspora, like Schoenberg. But quite simply he was detached from his cultural roots, something which has always made life hard for exiled Russian artists, whether they are like Chagall and resort to recurrent images of old Russia, or like Solzhenitsyn and more or less cease to work creatively.

Since the present series of Music Handbooks is younger cousin to an earlier (and continuing) series devoted exclusively to opera, it may seem perverse to include in it a study of a work which most enthusiasts – starting with the composer himself – have regarded as an opera. There is a reason for this but no real excuse. The reason, quite simply, is my own opportunism, backed up by the publishers' tolerance. Invited to contribute to the Cambridge Music Handbooks, I suggested Oedipus rex knowing perfectly well that I could defend the choice on the grounds of the work's notorious generic ambiguity, even though I fully intended to prove beyond doubt that that ambiguity is a myth. It is true, of course, that Oedipus has often been given in concert form, ever since its Paris première on a theatre stage but with no production and no sets or costumes beyond black drapes and sombre garments. Today it is perhaps more often staged than in the past; but it still more than survives as a concert work, and for this reason – if for no other – I hope it will not be thought of me that ‘scripsi quod nefastum est’.

Certainly, no excuse is needed for devoting a whole book to Oedipus. Of all Stravinsky's works outside the popular early ballets, it is perhaps the most widely admired and certainly one of the most frequently played.

Like most pigeon-holes in music history, the term ‘neo-classicism’ collects a lot of mail for quite different addressees. What looked at first like a bachelor flat turns out to be what T. S. Eliot's Prufrock would have called a one-night cheap hotel, where the clientéle is numerous, mobile, and often pseudonymous. The streets lead us to the overwhelming question that we cannot, unlike Prufrock, avoid asking: What is it?

For a start we can answer that it is, comprehensively, a misnomer. What the great aesthetic postmaster seems to mean by ‘neo-classicism’ is neither new (‘neo’) nor classical; or rather its classical elements are not new, and its new elements are not classical. There is new classicism all around us in music: in Berlioz, Mendelssohn, Tchaikovsky, Brahms – but also in Palestrina, Bach, Beethoven, Schoenberg and Panufnik. But it is not – or only rarely – ‘neo-classical’. These composers are, of course – by one of those delightfully Chestertonian confusions of the English language – classics, and their music is, by another such confusion, classical (in the sense that, for some reason, Irving Berlin's, Elvis Presley's and Miles Davis's is not). But its ‘classicism’, such as it is, has nothing to do with either attribute. It consists in an unconscious or semi-conscious desire on the part of the composer to ‘classicise’ his writing: to reject extravagance and obfuscation in favour of clarity, balance, restraint and a strict formal economy.

To turn from the external politics of the creation of Oedipus rex to the actual process of composition as shown by the composer's sketches and drafts is to follow Plato from the superficial world of shadow and illusion into the deeper reality of ultimate forms. This is a familiar experience to students of Stravinsky's sketchbooks. They reveal a mentality completely absorbed by the act of discovery, as described so vividly, a few years later, in the third Harvard lecture: ‘we grub about in expectation of our pleasure, guided by our scent, and suddenly we stumble against an unknown obstacle. It gives us a jolt, a shock, and this shock fecundates our creative power.’ Stravinsky nearly always grubbed at the piano, but whenever he encountered ‘something unexpected’, he would write it down.

The sketches are thus a record not so much of arduous labour, as with Beethoven or Mahler, as of moments of recognition or, to use Stravinsky's own term, ‘observation’. Page by page, the work in hand materialises on paper in the form of precise musical images that differ little in essence from their appearance in the finished score. Hardly anything goes unused. Once notated, an idea will be refined in various ways. From time to time, its elements may be arranged in slightly altered sequence; barring, articulation, verbal underlay and other secondary details may well be modified or added, melodic intervals may change.

Anyone who has ever tried to follow the plot of Stravinsky's Oedipus rex intelligently with no prior knowledge of the story will know that it is in fact an impossible task. Even for the dwindling tribe of Latin scholars, there is little hope of understanding exactly why, for example, Oedipus takes so long to realise his true situation as King of Thebes, for the simple reason that in compiling the text Cocteau seems deliberately to have left out such vital information. These omissions are usually attributed to snobbery. Stravinsky himself came round to some such opinion: ‘The line “And now you will hear the famous monologue, ‘The Divine Iokaste is dead’ ”, is intolerable snobbery. Famous to whom? And no monologue follows, but only a four-word singing telegram.’ But the sketchy nature of Cocteau's dramaturgy is closely bound up with the whole theatrical genre of the work, with its conscious use of an arcane language, its statuesque stage idiom, its allusive, tableau-like format. In a tangible sense, Oedipus rex is not only about the fate of a king called Oedipus; it is also about the ways in which such a profound but apparently remote tale can be given meaning to modern audiences in the modern theatre. As such, it is by no means isolated, either in Stravinsky's work or in the theatre of his day. So before examining the opera-oratorio itself, it will be worth looking at its context in the work of Cocteau and others.

After the three Munich performances, Mozart tried to pave the way for a Viennese production by performing parts of Idomeneo at concerts. But the end of the National Singspiel and the engagement of an opera buffa troupe effectively barred the way to Idomeneo at the Burgtheater, whether in Italian or, as Mozart probably intended, in German. The single 1786 performance was in a private theatre; the first public Vienna production, which was in German, came only in 1806.

There seem to have been no performances of Idomeneo during the 1790s, but Mozart's posthumously growing fame led to a spate of publications, including plans for issuing his complete works. Three vocal scores of Idomeneo were published in 1797-8, and in 1800 the overture appeared in the popular domestic form of a piano duet. The full score was first published by Simrock in 1805 (entitled Idomeneo ossia Ilia ed Idamante). Other vocal scores appeared throughout the century, and in 1868 Breitkopf and Härtel initiated a series of operatic full scores with Idomeneo. International interest is evident from the publication of vocal scores in London and Paris in 1811, not apparently related to performances.

Most productions were in German, the first in 1802 at Kassel. Idomeneo's French origins are reflected in a strange production in Paris, described as with ‘words by M. Caigniez, arranged from the opera by Mozart by J. L. Bertin’, in 1822. The cast included confidantes for Idamante and Ilia, but Elettra disappears, her first aria being incongruously transferred to Ilia (Ilione).

Mozart's connections with the Munich court dated back to just before his sixth birthday, when he was taken there to be exhibited: it was his first venture as a musician, probably his first of any kind, outside his native Salzburg. He was there again the following year, at the beginning of the lengthy tour that was to take him to Paris and London, and once more, on the way back, three and a half years later. The fourth visit was in 1774-5, when his opera buff a La finta giardiniera was given at the Salvator Theatre during the carnival season; anxious to make an impression in Munich beyond the opera house, he wrote a number of liturgical works, among them a powerful Misericordias Domini (K. 222/2O5a) in a learned, contrapuntal style.

The years 1777-8 saw Mozart's crucial journey to Mannheim and Paris, seeking career opportunities away from the provincialism and pettiness of Salzburg and its court. Munich was his first important stop. In a letter home to his father (2 October 1777) he fantasised about staying there and earning a living composing operas:

He added that a German serious opera was in prospect ‘and they are very anxious that I should compose it’.

The libretto

For all its allegiance to Franco-Italian reform opera, Idomeneo remains within the eighteenth-century norm of alternation of action and introspection, recitative and aria. The three ensembles are more closely merged with the preceding recitative than most of the arias, and all proceed from dialogue to combined voices, yet they function like multi-voiced arias more than the developing ensembles Mozart cultivated in opera buffa; they do not further the action. Three static numbers involve solo and chorus, and the full choruses are largely decorative; only the shipwreck, the end of Act II and - static as it is - No. 24 (‘O voto tremendo’) involve the chorus closely with the principals.

The general structure of the libretto is admirably simple; all three acts begin intimately, and end in public scenes. Act I unfolds the pattern in two stages. A monologue and dialogue are followed by a semi-public scene of rejoicing. The first dramatic event is the announcement of Idomeneo's shipwreck. We are then confronted with the intimate feelings of Elettra, and, after the storm, with those of Idomeneo, before another dialogue. Haltingly, the action returns to soliloquy (Idamante) before the public scene of the divertissement. Such a stop–go process is perhaps a necessary concomitant of exposition. Mozart's cutting axe hardly fell on Act I, which lasts fifty to fifty-five minutes.

The second act, the shortest at around forty-five minutes, begins with an opera seria sequence of dialogues and soliloquies before the public scene focuses on Elettra. After the trio the finale brings the elements, the people, and the king into fierce dispute.