What on earth can Britten have meant? Was he referring to residual infelicities in Montague Slater's endlessly worked-over libretto? Did he now feel that some of the more picturesque numbers, such as the "Goodnight, Vicar" ensemble in Act 3, held up the musico-dramatic pacing? Had he even become worried about a certain ambiguity in the conception of the protagonist, as implied by his praise of the voice of a young Canadian Grimes in 1950: "It was not too heavy, which makes the character simply a sadist, nor was it too lyric, which makes it a boring opera about a sentimental poet manque."
Or indeed, vulnerable as ever to criticism, had he half come to believe the recurrent strictures from less sympathetic quarters which seemed to greet each of his operas: that his style was derivative, his textures thin and his word-setting mannered; that his handling of denouements was somehow inhibited; that instead of concerning himself with romantic love, he seemed obsessed by the corruption or destruction of innocence?
Yet next Wednesday's 50th anniversary of the first performance of Peter Grimes at Sadler's Wells on 7 June 1945 is being widely celebrated; Billy Budd and A Midsummer Night's Dream are currently running at Covent Garden and ENO respectively; and a glance through the schedules of the world's opera houses, to say nothing of the current record catalogue, simply confirms the realisation that Britten remains the only composer since the last war to establish an entire operatic output in the international repertoire comparable to those of Janacek, Richard Strauss and Puccini earlier this century, and of Wagner and Verdi in the last. And since none of the persistent criticisms, however one-sided, is without just a tiny bit of truth, the question remains: how did he do it?
Aaron Copland remembered once asking Britten what he thought was the essential requisite in composing operas - expecting to be told sense of drama, swiftness of scenic evocation, or something of the sort. But Britten replied, "The ability to write many kinds of music."
If one surveys even the first half-hour of Grimes, it is clear what he meant. The music progresses from jagged recitative, through ostinato-based crowd textures, to a long-line vocal duet; then, by way of an orchestral tone poem, to a sustained tableau with choral refrain and more volatile interpolations of action music, culminating in a hieratic aria for the soprano and a violently contrasted storm chorus for the whole company. So it goes on.
And if one considers Britten's operatic output as a whole in the light of this mastery of varied means, then a rather remarkable pattern emerges. There are, in fact, three quite distinct grand operas for the largest forces: the traditionally narrative Grimes (1945); the darkly allegoric Billy Budd (1951), projected as if through the memory of Captain Vere; and the emblematic, pageant-like Gloriana (1953). Then there are three works for medium-size, as it were Opera-Comique or Volksoper, resources: the neatly turned triple comedy A Midsummer Night's Dream (1960); the televisual morality Owen Wingrave (1970); and that last, fluidly structured meditation on order and chaos, Death in Venice (1973). Then, yet again, for companies comprising a small team of singers and a dozen-odd players, there are the three chamber operas: the formalistic tragedy of The Rape of Lucretia (1946); the vernacular burlesque Albert Herring (1947); and that obsessively cumulative mystery The Turn of the Screw (1954).
As if such a triad of trios were not enough, there is also, for still smaller groups, the trilogy of church parables Britten conjured from such pre-operatic traditions as Noh theatre and medieval music drama, but which served as such a stimulus to the development of music-theatre in the 1960s: Curlew River (1964), The Burning Fiery Furnace (1966) and The Prodigal Son (1968).
If he had succumbed to inducements in the late 1950s to write a bona fide musical, this could have completed yet another trio of singspiel- like pieces, inaugurated by the American operetta Paul Bunyan (1941) and continued by his arrangement of that echt-English 18th-century musical, The Beggar's Opera (1948). And not least, as if to introduce future generations to his operatic world, there were Britten's three stage works for children: the vastly tuneful tale of The Little Sweep (1949); the touchingly communal Noye's Fludde (1957); and the shorter vaudeville composed for the Vienna Boys Choir, The Golden Vanity (1966).
Even if one accepts Oliver Knussen's submission that an almost chess- masterly ability to organise heterogeneous possibilities lay at the heart of Britten's art, it seems unlikely that this extraordinary symmetry of operatic output could have been planned from the start. And, of course, certain might-have-beens would radically have altered the picture: The Tale of Mr Todd, which the publishers of Beatrix Potter refused to allow; the King Lear project of the mid-1960s too quickly leaked to the press; and the setting of Anna Karenina with Galina Vishnevskaya jointly planned for Covent Garden and the Bolshoi but stymied by the Soviet invasion of Czechoslovakia in 1968.
But what a simple categorising of the output in all its variousness does bring home is the foresight and realism of Britten's career. Scarcely was Grimes triumphantly launched than he concluded that concentrating entirely on a sequence of similar grand opera projects - after the tradition of Puccini, Strauss or even Janacek - was no longer either practical or sufficient in the post-war world. One could argue that a large part of his genius lay in the flexibility and skill with which he proceeded to turn almost any venue and collection of performers into opera. Not one of his contemporaries was remotely as versatile.
In the face of this achievement, even the most criticised shortcomings seem by now more in the nature of defining limits. One hardly goes to Britten for operatic textures of a Straussian opulence, socio-political satire of a Weill-like bite, or avant-garde experimentalism of a Berio- type elaboration. Nor, mindful of the need to sustain the plausibility of singing as a dramatic convention, did he ever attempt a libretto set in the everyday language of the present time - unlike Tippett's four quixotic attempts to do just that.
Yet this in turn hints at perhaps his most striking achievement. One may admire the many innovations of operatic structure and stagecraft he evolved over his career; one may marvel at the range of pace, character and atmosphere he drew from his accompanying orchestras and ensembles; one may brood endlessly over the psycho-symbolic implications of his dramatic themes. But, as in the operas of the predecessor he revered most, Verdi, the focus first and last remains the voice; surveying the 17 stage works, one can find scarcely a bar where he allows orchestra or action to overwhelm his singers. Britten was not a canary-fancier: "The singers must, of course, have good voices, but these should be used to interpret the music, not for self-gratification." All the same, one suspects he would have agreed with Maw's submission that, in the last resort, opera is actually about singing.
n Britten operas on Radio 3: A Midsummer Night's Dream, tomorrow 6.30pm: Billy Budd, Mon 7.25pm; Peter Grimes, Wed 7pmReuse content