CLASSICAL MUSIC / Finished business: In the light of his definitive new study, Brian Newbould looks back on his experiences of completing Schubert symphonies

SITTING in the Parry Room Library at the Royal College of Music 15 years ago, turning the pages of the autograph sketch of Schubert's unfinished Seventh Symphony, I was intoxicated by a cocktail of emotions. First came the elation of savouring a tantalising outline of an entire, sizeable and unknown symphony by one of the great symphonic composers. Then there was thankfulness that when George Grove came into possession of the sketch, left it on a train in south London and had to live with the guilt for a night, it turned up the next day at Norwood Junction.

Offsetting my excitement at the symphonic sweep of such a racily (and therefore sparingly) notated score, came the doubts: whether I, or anyone else, could assume and justify the responsibility of representing its creator's half-realised thoughts to posterity. Moreover, Brahms's dictum echoed menacingly through one's conscience: '. . . please make sure that no indecency is perpetrated with it]'

Mendelssohn saw Schubert's sketch and would not finish it. Brahms, having not seen it himself, took Mendelssohn's restraint as a reason to stay the hand of any other who might venture to do so. So was it anything but folly to rush in where they feared to tread? Mosco Carner, writing in 1946, typifies the conflict between moral principle and human curiosity: 'Though on principle an opponent of such undertakings, I would welcome an occasional performance. . .'

Carner cannot have it both ways. It was the wish to make Schubert's skeletal thoughts available to the ear that drove my six-month full-time tussle with the sketch. My objective was a student performance before a small local audience. For the sequel, blame (or thank, as you will) the BBC. It was at the first broadcast performance, by the BBC Symphony Orchestra at Cheltenham in 1978, that I was prompted to reflect on the peculiar difficulties that would baulk a composer who might try to finish this symphony. Both Iain Hamilton, who had a premiere in the same concert, and Lennox Berkeley, who was also present, denied that they could have completed the work. This was not so much excessive modesty, though there was a touch of that, as a recognition that one composer would have to shed his own inventive persona to assume that of another. My own credentials as a composer - a roster of works that could be counted on the fingers of two hands - hardly brought me up against this problem.

The subsequent currency of my realisation of the Seventh was as unpredicted as the appearance, just after its completion, of a sketch newly identified as being a symphonic attempt of Schubert's last months. The problems that this 'piano sketch' posed were different from those of the Seventh's orchestral sketch, and on balance more formidable. I doubt if I could have undertaken a completion of this Tenth Symphony without the experience and confidence gained in fleshing out the earlier one. Again, the motive was to share with others music of immense interest - and sometimes profundity - which would otherwise go unheard.

How do I see these realisations 10 or more years on? Much careful weighing of detail went into making a performing version of the Seventh, but this was compatible with the other prime obligation to Schubert of making the symphony 'go' as a concert piece. Whatever weaknesses may remain since the last round of adjustments made during 15 hours of recording sessions in Berlin in 1980, I find it works well for me, particularly in the concert hall, and the slow movement remains one of Schubert's loveliest. My love of the work, I confess, surpasses that for Schubert's Sixth, for the Seventh represents a timely enlarging of his vision, despite its non-consummation.

As for the Tenth, I wish there were some way of knowing what structure was in Schubert's mind for the first movement. Yet, distanced from decisions which took up to two years to make, and with perceptions honed by contact with countless informed listeners and university audiences at home and abroad, I have not yet come across solutions which better command my confidence. For the sake of the soul-searing wonders of the slow movement, whose consoling last major-key theme seems to haunt all who hear it, not to mention the brave experiment of the scherzo-finale, I cannot feel that the audacity to speculate was misplaced.

Completing the 'Unfinished' itself was a relatively minor mission, even though it was the only one to necessitate a span of pure composition, with not even a single line on which to build in one section of the third movement. For many it is not an unfinished symphony so much as a finished half-symphony, and, given the picture of unsurpassed sublimity it presents, how dare anyone come along and stretch the frame? I would perhaps not have thus risked eternal damnation had not Philips invited me to do so in order to make consistent their planned Schubert album.

It is hard for all of us to let go of our long-held perceptions of this half-symphony and to get to know the completion as well as we know the familiar torso, so as to be in a position to make valid judgements. At the same time, many who listen, some critics not excepted, find it hard to separate separable issues. Did Schubert first compose the Rosamunde Entracte as the finale to the symphony? Is it a good finale? Is it worth attempting a 'reconstruction' in order to see how Schubert was thinking of proceeding after his second movement? - and so as to hear the familiar first two movements in a symphonic context of some sort?

The benefits of following the fortunes of these 'new' symphonies over the years have been several. I have learnt much about conductors. One made several changes to minutiae in my score because they were 'not Schubertian', yet most of those changes were to strands of the final product that were Schubert's, not mine. Another saturated himself in the Tenth Symphony for a year, studied the sketch in Vienna, and came to discuss the work a few weeks before his performances and radio recording of it.

I have had enriching debates in locations near and far, discovering dedicated Schubertians in distant lands. And I have begun to glimpse, with the recent QEH performance and Virgin recording of No 8 on period instruments, by Mackerras and the Orchestra of the Age of Enlightenment, how such an approach may require a review of the odd detail of instrumental usage in the realisations. Precise intonation in the woodwind proved elusive at one point, for example, and we tried an alternative arrangement of notes. Would Schubert have avoided my original contour here, or tolerated the imprecision?

Further than that, I have now been able to make a fresh study of Schubert's finished symphonies in the light of the unfinished ones. These studies now lead me to reconsider other parts of the vast Schubert oeuvre. One result must be that the old view of Schubert as an intuitive musical innocent who penned a ditty on a cafe menu between coffees and had no use for Beethoven's metaphorical chisel will eventually die. One cannot witness the manifold reshaping of themes and of contrapuntal working- out revealed in the sketch for the last symphonic movement he envisaged, or discover in his opera Die Zauberharfe the only full-blown musical palindrome of the nineteenth century and the most daring one ever ventured within the tonal harmonic system, and not see him as a musical intellectual as sophisticated as the next and more technically adroit than most.

Brian Newbould is professor of music at the University of Hull. His book Schubert and the Symphony: A New Perspective is published by Toccata Press ( pounds 25)