I've just been reading an interesting exchange in the music blogosphere between critic A. C. Douglas (Sounds and Fury) and composer Matthew Guerrieri (Soho the Dog) The point of contention is a post by Guerrieri discussing his systematic approach to composition; for him, composition is less a matter of sudden inspiration than of methodically testing a variety of musical possibilities on each material. Douglas takes umbrage at this; for him, successful compositions are based on fully developed ideas which are worked out to create a continuous narrative, and any compositional artifices should be carefully concealed. In fact, Douglas views this emphasis on process to be a major flaw in contemporary composition. The workings of the machine should be invisible; you judge the product that comes out.
This disagreement is interesting for me only in that it reveals how differently I and ACD listen to music. For example, Tippett's Concerto for Orchestra, a work I admire, is based on an obvious compositional conceit; the orchestra is divided into small chamber groups (flute and harp, tuba and piano) which state musical material idiomatic to themselves. These musical ideas are then overlaid in different ways. The system is strict; the flute never plays the themes entrusted to the clarinets, for example. Once he's tried all the superimpositions he wants to, the movement is over. There are plenty of other examples of twentieth-century pieces which I love not despite, but because of the fact that their compositional conceits are obvious. Think of Stravinsky riffing on Mozart operas in The Rake's Progress, or Ives wondering how you could musically reproduce two different bands playing totally different pieces simultaneously.
In general, the music I dislike from the past century leaves me with no clues as to how it's put together. To take another Tippett example, I'm sure The Vision of St. Augustine is really cleverly written and an analyst would have a field day with it, but after struggling through it with a score a few times I find myself incapable of perceiving any musical coherence in it whatsoever. But the works which really infuriate me in this regard are the works of Brian Ferneyhough. After listening to a CD of his chamber works, I can't say that I understood what was going on for more than about ten seconds at a time. Occasionally I'd catch a hint of a phrase I could latch on to, a recognizable melodic contour, or something. Then, chaos. For minutes at a time. It's so rare that I feel like that when listening to music that I was actually horrified. I went to the library and consulted a volume of his prose writings, which I found equally impenetrable. You get the sense that, far from making the nuts and bolts of his composing (and prose!) visible, he's cultivated a surface layer of noise to scare off all but the initiates.
I'd bet I'm not the only one who shares this approach to music; when you can understand some of the nuts and bolts behind the piece, even in the anecdotal ways I've just described, it starts to make sense. Rather than being opposed to Douglas's ideal of a "seamlessly coherent musical narrative", these procedures can help bring the narrative into the foreground. The atonal pitch structures of Tippett are readily coherent when the form makes sense (the Concerto) and difficult to penetrate when the form is not audible (the Vision). Remember, the critics of twelve-tone music never complain that its compositional conceits are too obvious; they complain, instead, that its procedures are too complex to be perceived on an audible level. So let's have more fugues, and fewer pieces with titles like Unity Capsule.
The above should not be interpreted as one-sided criticism of Mr. Ferneyhough; the fault may well be mine, and not his.