Jean-Jacques Nattiez's book Music and Discourse: Toward a Semiology of Music (available from Powells or Amazon) hypothesizes a musical work as a series of configurations, from composition to listening, in one of the most stimulating works I've read in some time. It explicitly acknowledges a problem with music analysis that I have always felt intuitively, which is that very often the explanations that we get for new music don't seem very relevant toward understanding. How many times have I read that such-and-such a piece is based on a specific tone row? Am I supposed to hear the row as a 'theme' in the way that Beethoven uses the technique? Not likely. Am I supposed to hear inversions (retrograde and otherwise), or uses of the row in chords? Not very likely either. One thinks of the detailed documentation that Lev Koblyakov made of the compositional processes involved in Pierre Boulez' work Le marteau sans maître, which are not required to appreciate the beauty of the piece.
Instead of considering that an account of the compositional processes explains a piece of music, Nattiez distinguishes three distinct areas of concern in music analysis, what he calls the tripartite analysis: the process by which the music is composed, the musical work itself (which Nattiez identifies with the score because a realization of the score is already a kind of interpretation or analysis), and the process by which the music is perceived. Of course, being French, he must cultivate his jargon and create new words for the compositional and perceptual processes, but this is the cost of new ways of understanding. And my oversimplified explanation should not eclipse the nuances and subtleties that Nattiez evinces in his descriptions of the musical work.
One of the most fruitful suggestions from the separation of concerns for Nattiez' tripartite analysis is that multiple analyses are possible for a specific situation, and Nattiez spends some time in the final chapter covering some of the many analyses of Wagner's Tristan chord. Although Nattiez refuses to go to a complete relativism that would allow any possible explanation, he seeks to place an analysis in a larger schema that he calls a 'plot', a pre-established set of criteria that an analyst will use for guidance. Musicians perform the same operation. When I play Takemitsu's piano pieces, I have in mind the whole story that I have constructed in my mind for Takemitsu's place in music, and which I've outlined in previous posts. When I read Burt on Takemitsu, I may gain understanding about the compositional processes, but it is the perceptions that I have about Takemitsu that are more important. I don't hear the scale that he may have used in a particular composition, I hear the poetic effects and the colors.
Nattiez leaves out one very important process in musical understanding, which is the work a musician does in order to turn the musical score into an audible artifact. Traditional analysis doesn't necessarily translate into better performances, but often seems directed at creating better composers. A performer must not only make decisions about the meaning and intent of a musical work, but must be able to translate those decisions into sound. This is a different kind of work altogether from a scholarly analysis. Nattiez also glides fairly quickly over contemporary electronic works that exist only as recordings, saying that it 'would be easy enough to identify and describe the sound-objects that make up these works' — as a reviewer of various electronic works, let me respectfully disagree. An impartial description of an electronic recording is extraordinarily difficult if not impossible to make, and the examples that Nattiez adduces (such as the listening score for Ligeti's Artikulation) were all made with significant assistance and cooperation from the composer.
While I was reading Music and Discourse, the John Cage mailing list had a discussion about Atlas Eclipticalis, a work I've written about before. Various correspondents testified to a significant oral tradition around Cage's music that remains essentially undocumented. Daniel Wolf paraphrased composer Richard Winslow: "if you want to reproduce something precisely, transmit it orally; if you want to guarantee that something changes over time, write it down." While I somewhat reluctantly agree with Petr Kotik's assertion that all notated music must be learned by direct encounter (after all, I still take piano lessons), Cage's scores are tantalizing entry points to a unique perspective on sound and music. And frankly, it's not easy to find a teacher with a direct lineage back to Cage, in the way that my previous teacher traced her Bach lineage to Rosalyn Tureck. The ambiguity inherent in a Cage score is part of the attraction, and I feel compelled to deal with his music, and that of his successors, because even when I get it wrong, the effort still moves me closer to understanding and enlightenment. Nattiez illustrates some of the problems inherent in interpreting a score, the decisions a musician must make to realize music from the score, all while examining the assumptions that provide the unstated background for this sublime and mysterious activity.