If a composer is lucky enough to have a publisher who is willing to finance a revised score after publication of the original, it is reasonable to take advantage of the offer. I have done this only twice because I am always too busy with the next enterprise. Besides, a work represents what you did at a particular point in your life. It is commonly stated that, ‘A work of art is never completed; it is merely abandoned.’ There is always something that could be added or improved. This defines why I am reluctant to make revisions, because I am never satisfied and would never stop revising. I put what I learn into the next composition. Revising the details of a score does not necessarily create a definitive solution. Each performance requires adjustments to balance, dynamics and tempo, according to the acoustic of the arena or recording studio. No score can be absolute in terms of such details. This is even more of an issue when real-time electronics are involved, and one of my own works of this kind illustrates how this can create the need for post-première revision. At the Still Point of the Turning World was composed in analogue days, when human beings controlled the electronics in performance. The sound source was a solo oboe leading to six tape delays ranging from 3.2 seconds to a minute, each with amplitude control, ring modulation and controlled feedback capabilities. These treatments were applied by 12 potentiometers under the control of 12 musicians. Together with the oboist, they performed with the conductor, who acted as a metronome for the exact co-ordination of the resultant ‘orchestration’. The score for this original edition would seem very primitive to twenty-first-century musicians. With the digital program devised by Lawrence Casserley, the score has been revised to suit a modern performance with only the oboist as performer, with the aid of a click track. There are three commercial recordings of this version (including my own) which illustrate very distinctive interpretations, but all are true to the revisions made to the original concept. I am quite nostalgic about analogue days!
My reluctance to revise works is not commonly shared. Many composers more distinguished than me have indulged in revisions after the première with justification. Boulez rewrote the improvisatory elements of Don (from Pli selon pli) in a definitive version for the second edition (1989) because the original version was too complex for the conductor in its graphic structure. But it is essential for a conductor to study the first version in order to comprehend the nature of the improvisatory idea. The second version looks comparatively bland. The second edition of Le marteau sans maître (1957) has drastic revisions of metronome marks and includes graphic symbols for the conductor; there are no radical changes to the notation, however. The decision to revise the notation of Don might have been made by Boulez in recognition of the costly rehearsal time required to rearrange the seating of the orchestra, to explain how the players should respond to the complex conducting gestures and to allow the orchestra to digest and learn to react correctly to the information and its musical implications. In the belief that the first version represented the essential nature of the improvisatory concept, I took advantage of having substantial rehearsal time available with students of the Royal College of Music and prepared a performance of it – with the added luxury of the late Jane Manning singing the vocal part. It revealed the colour and intricacy of the polyphony vividly. The performance was recorded and remains in the archives of the Royal College of Music.
Birtwistle was persuaded to rewrite his pulse piece Silbury Air because of the beating complexities of the original, which is based on a series of pulse labyrinths establishing a framework of metric modulations, where time signatures indicate the varying subdivisions of a unit while integers define the number of beats in each bar. It is not easy to conduct the original version, but it is a brilliant concept which the graphics convey exactly. The second edition (2003) reverts to conventional notation with revisions to the orchestration and recomposition of some sections. The logic of the original peeps through in a 25/16 time signature at one point. My conviction is that the first edition conveys the spirit of the music more vividly because of the demands on the conductor, who should find a way to fulfil the intentions of the composer no matter how complex.Footnote 91
The need to revise a work after the first performance is clearly illustrated in Stravinsky’s search for the best solution to barring the irregular pulses of Le sacre du printemps. In his sketches, a principal motif is written as 5/8 5/8 6/8 5/8. In the 1947 edition it is lengthened and changed to 5/8 5/8 3/8 4/8 5/8 6/8 5/8. The Édition Russe version of 1921 opens the Danse sacrale with 3/16 5/16 3/16 4/16, whereas the 1947 Boosey & Hawkes edition has 3/16 2/16 3/16 3/16 2/8. Had Stravinsky been a more accomplished conductor the solution might have been resolved in an earlier edition. But the innovation was historically shattering, showing how a modification of original intentions can be enhanced by the trial-and-error process in confirming a concept of revision. Perhaps I should spend some time revising and improving my own scores. One I refuse to change is in a section of my Concerto for Orchestra where the conductor is required to conduct five with the right hand against four in the left. It is essential to the counterpoint. The conductor has to do some practice.
While I am reluctant to pursue compositional revisions to a work after a première, the issue of collaboration with a soloist has an extra dimension. This requires consultation before the composing commences, especially if the composer thinks of the work as a sound portrait of the soloist. Application of experimental issues discussed will leave uncertainties in applying them to the composition, especially if the composer does not play the instrument. The première then becomes a necessary element of the compositional process, as afterwards the soloist will offer final recommendations with regard to the details involved. But my stance on leaving the material of the original composition unaltered remains.
Revisions which involve cuts are not uncommon. The duration of Rachmaninov’s Second Symphony in its first performance was almost an hour. The composer devised cuts which reduced it to 35 minutes in subsequent performances. It is to their credit that orchestras have returned to the original score in recent years. In the case of Michael Finnissy’s Piano Concerto no. 1, the work was completely rewritten at half the length of the original version and premièred in this new form by Ian Pace in 1996. Pace describes the new version as ‘a distillation of the earlier work’.Footnote 92 The desire to distil the substance of a work into a more succinct statement contrasts with Boulez’s need not only to recompose some works, but also to lengthen them after a première: examples include Éclat and Figures–Doubles–Prismes. Boulez’s method of composition makes the material infinite in its compositional equations. The same could be said of the variation principle in diatonic music: Bach could have composed hundreds of variations for Johann Gottlieb Goldberg, and the existing ones explore the material with unceasing inventiveness.
Conductors sometimes have to deal with problems in works which do need revisions after a première but fail to be emended by the composer. In spite of such problems, the quality of the music can be of a very high standard. The Winter of the World, an ensemble piece by Elisabeth Lutyens, is such a case. The problem was a very common one in music of the 1970s and 1980s, when improvisatory elements were fashionable. The desire for motivic freedom often resulted in the abandonment of the barline and the time signature. Works such as Berio’s Sequenza VII, Maderna’s Giardino religioso and Lutosławski’s Preludes and Fugue are representative of this characteristic. Being such a wonderfully eccentric person, Lutyens presented her score with, at one point, 40 arrows indicating demarcation points within one bar. Players waiting for the cue for arrow 17 would already be lost. The prospect of waiting until after the première to make revisions was not an option. As a conductor, I inserted barlines where possible and numbered the arrows in relation to the beating patterns I prescribed (see Figure 1). These graphics remained for use in subsequent performances. As in other scores which I have discussed, the impracticalities were induced by the ideas which informed the work. It is part of a conductor’s task to interpret them with solutions which project the idea successfully. The post-première revisions are then an asset to the performers. My contention is that such changes should never preclude an awareness by the performers of the original score.
Figure 1 Elisabeth Lutyens, The Winter of the World (1974), figures 38–9.
