This (apart from, ‘is there any money in it?’ and ‘what kind of music do you write?’) is the question a composer is most often asked. And it is an interesting one, because more complicated than the asker often realises. When I first listened to music, the question which fascinated me was how the composer managed to hear the different lines and instruments all at once. And this is the unique aspect of western music: no other music shows such a degree of simultaneous unfolding: bass and soprano, woodwinds and strings, a three against a four. Then I studied harmony and counterpoint, and the process did not seem so magical any more. Knowledge abolishes mystery. But harmony and counterpoint provide rules that govern the period of general practice from the baroque to the romantic. What about the modern period where such rules no longer apply? A modern work like Stockhausen’s Gruppen deploys three orchestras playing different music in different tempi in an astonishingly complex interplay. How much of this web of sound did Stockhausen ‘hear inside his head?’ Little, I suspect. Does this make the composer or the work a fraud?
This brings us to the definition of the word ‘hear’ in the question. A composer hears the score in the sense that he or she assumes responsibility for what he has put on paper. This means the entire web of connections – of pitch, rhythm, colour – that makes up the work. In this context, to expect him to ‘hear every note’ is a bit like expecting an aeronautical engineer to push the aircraft into the air by willpower. This is not his job. His job is to set in motion a series of actions that result in flight. This analogy defines the composer’s relation to his score, and the sense in which he hears it.
And yet there is one aspect of composing where the resulting sound remains largely speculative: orchestration. The reasons for this are practical: the composer cannot hire an orchestra to try things out as he goes along. (The last composer to enjoy that luxury was Haydn at Esterhazy.) So he has to put that unusual doubling (oboe and marimba) on paper as an act of faith and hope for the best when it is played.
Composers can be guilty of many sins that recordings disguise. I remember my disappointment when I heard my first live performance of Berg’s Violin Concerto. The violin for long stretches was inaudible. This may have been the fault of the performance, but I suspect not. I had come to know the work through recordings that disguised miscalculations in the orchestral balance by making adjustments in the studio.
A radically new element has entered the picture with the invention of computer software and midi playback. A composer may not be able to hire an orchestra, but he can own a studio where every pitch and rhythm is played with absolute accuracy. Surely this is the utopia we have all been waiting for. It has not proven to be so. But firstly let us admit the amazing advantages. You can hear how the most complex rhythms actually sound, freely devising polyrhythms that make Conlon Nancarrow sound like nursery rhymes. (The end of my brief infatuation with this composer coincided with my first experiments with midi playback.) And counterpoint of any complexity is no obstacle to a machine. The young Stravinsky badgering his children to take a voice at the piano which his fingers could not reach would have been spared much heartache, as would the children. So what is the problem? The problem is that the sound of midi playback inhabits a lifeless, two-dimensional space. Trying to work out what your piece sounds like from it is like trying to appreciate a Shakespeare monologue that is transmitted by an astronaut from the space shuttle. I am reminded of Truman Capote’s brilliant put-down of Jack Kerouac: this isn’t writing, it’s typing. For me, Sibelius software remains a form of typing and not the ‘compositional tool’ that many people claim it is. Orchestral and chamber music especially need to attain a point of density before they ‘sound’ (vocal music, by contrast, can thrive amazingly on a calculated parsimony). And speaking for myself, I can achieve that density only with pen and paper.
The composer ‘hears the piece in his head’ in the sense that he carries a platonic ideal of what it sounds like. No physical execution of the piece, however good, will ever approximate to this. (This explains why composers and performers generally experience recordings of their work as a form of torture.) A composer’s life is spent in narrowing the gap between these two: the ideal and the real. He will never succeed, and this is just as well, because then he would stop composing.