The Music Of Speech and the Music of Machines

There is another layer of complexity to this that we’ve got to consider. Music seems to have two sides: a lyrical side that is related to speech, to talking, to the human voice, and a physical motion side, that is related to dance, to repeating motions, to our heartbeats and pulses, to work over time, to machines.

In a simplistic way, we can think of these sides as a sort of yin/yang of melody and rhythm, patterns of pitch and patterns of time. They usually exist simultaneously and interactively, and often are used in tension against each other. In the past century, particularly, the rhythmic side of music has developed strongly, in mimicry of the machines we work and play with, love and hate, in mimicry of the big clockwork society we live in and the mechanistic organization of our lives. A large part of our musical universe includes “machine” music, music that is essentially the mechanical working out of patterns, invoking the part of our mindplay that delights in watching things work, just like we watch those balls run down the tracks in those silly mechanical sculptures in our more cultural airports.

MIDI, and the current crop of sequencers, are ideal for such machine music. They are predisposed to the repeating loop, the layering of loops, and the working out of patterns. They predispose us to the creation of elaborate musical grooves, over which we make up yet more melodies with words about love and sex (Yeah, Yeah, Yeah! Woe, Woe, Woe!). Such a predisposition is neither bad nor good, it just IS. Some really nice music has been created this way, and it can be a fast, creative and exciting way to compose. Mechanistic formulas for the creation of music have a rich and proud history – the Baroque fugue is about as mechanistic as you can get, for instance.

Whatever. “Machine music” these days tends toward strong, driving rhythms and high energy generated by the cumulative effect of repetitive patterns, usually layered in increasing complexity as the music progresses. Such processes are at the center of much “techno” music, and when such music is criticized, the arguments against it are usually arguments against its mechanistic and repetitive character, its lack of “soul,” “humanism” and lyrical feeling. MIDI and computers get a bad rap because they are used to create “machine music” and may even have a predisposition to its creation. However, machine music is made by humans, not computers, and criticism directed at the tools are misdirected. If the music lacks soul, it is probably due to the composer and performers, not the instruments!

Aside from the question of the merit of “machine music” (I happen to like it - one of my favorite pieces of machine music is Stravinsky’s Rites of Spring), here is a more central question that pertains to MIDI. Is it really limited to such music? The MIDI spec seems to offer enough flexibility in terms of time and timbre control so that recordings of MIDI data gathered from live performances feel perfectly live. Fooling around with sequencers and live performances, and listening to the effects of quantization have led all of us to the insight that rigid quantization of MIDI data to the nearest eighth note usually trashes the musicality of such data. Given that, the answer is clearly no – MIDI is not limited to crude, mechanistic performance qualities. We arrive at such qualities in our efforts to simplify the MIDI data, to reduce it to something resembling the notes or dots on a page!

The Relationship Between Musical Performance and Musical Notation

Here is where it is useful to reconsider the relationship between “the music” and the music. Those dots on a page are the equivalent of a rigidly quantized MIDI score. If we play the dots “perfectly correctly,” which is to say exactly as they are written, then we will get as dorky a performance as the similarly instructed MIDI sequencer gives us. In both cases, it is bad musicianship. In the case of the MIDI performance, we blame the sequencer, even though it is our own bad judgment that leads to the dorky performance.

This should lead to an important insight. Real music is NOT made up of eighth notes and the like, and if you crap around with such things, you are going to need serious performance help to get some real music out of them. Good performers look at the notes and intuitively guess at the music probably residing in the heart of the guy/gal who made ‘em up and they just play it. This is called phrasing, expressiveness, musicality, etc. It seems to happen effortlessly, magically. In fact, the performer is creating the music, music that happens to be compatible with the printed instructions given to him/her.

This art of performance is a rich, powerful and deep one. It is an art that doesn’t lie in the notes, but in the transitions between the notes, in the patterns of change in the notes, and the patterns of patterns of change. And this is where our view that notes are the fundamental elements of music fouls us up. Music isn’t about notes, but rather about getting from one note to the next and from one phrase to the next. Performers do this intuitively in the terms of music-as-speech and music-as-motion I mentioned above. Meanwhile, because we were thinking about sound-as-notes, we created a synthesis/MIDI system designed to deal with just the notes themselves, not the transitions between them or the micro and macro patterns of timbre, articulation, timing and pitch that constitute musical phrasing. Damn. Makes it a little harder to get synthesizers, computers and MIDI language to play musically. It’s interesting to note that after Moog sold his company to Norlin, he went into the business of trying to create more expressive and physical controllers, because he perceived that was where the real development problems for synthesizers lay.

Post a Comment