Sunday, September 2, 2007

anecdotes rhythmiques

Several past posts by Kyle Gann (also here) reminded me about how much I've given up rhythmically to make my pieces playable by the performers available to me and how much I've allowed myself to be constrained by what others thought. I have to say that reading his blog in general has both shamed and inspired me and so I'm going to repeat and extend my comments here.

I was in an art rock band in the early 80s where a few of the tunes had multiple riffs of different, usually relatively prime, lengths played by different instruments. In fact, one number consisted only of a set of ostinati, one per part, and was titled 4.7 x 10^6 (pronounced 4.7 million), which referred to the approximate overall period of the whole mess in beats. I didn't use the technique that much in my own music, even though I may have wanted to, mostly - and I'm a little embarrassed to admit this - because I had read the annotation of example 35 in Messiaen's Technique de mon language musical where he describes "Our first essay in polyrhythm, the simplest, the most childish, will be the superposition of two rhythms of unequal length, repeated until the return of the combination of departure." The added emphasis is my own. There was something about that 'childish' comment that put me off the whole thing. Ach Gott in Himmel why do I listen to other people?

I was working for Yamaha back in the late 80s when the Finale notation program came out. Both Guy Garnett and I tried it and both had problems with the very first things we tried, and both for the same reason - its inability to handle partial tuplets, i.e., a tuplet which doesn't last for the entire length of time implied by its denominator, e.g., a single triplet quarter note. I was using such things in my postminimalist numbers and he was using them in his Stefan Wolpe inspired tunes. However, I didn't learn the obvious lesson from this: using this program is evil and will simplify the music you write. And so it goes...

I used to write a lot of meters with fractional bits, e.g., 4 1/2 beats of 4, which was clearly the correct notation. The music was supposed to sound like a little bit was dropped off the end - of a normal 5/4 measure in the above case - but I couldn't get conductors to beat it that way, even though it seemed really straightforward to me. They all wanted to make it 2/8 + 2/8 + 2/8 + 3/8, which is of course the same length but hardly the same feel. I think the hiccuping rhythms in the Concord Sonata may have been the first thing to make me think about using such things, but didn't Led Zeppelin take these kind of rhythms to the masses in the 70s? In what universe do classical players grow up? And why have I allowed them to browbeat me into their way of thinking?

I first started writing electronic music in the late 70s. I built with my own hands a small MIDI interface for a North Star S-100 bus computer as well as an 8-bit D/A and I wrote both a small polyrhythmically-oriented score description language to MIDI converter and a MUSIC-N like programming language for it in UCSB Pascal. It was unbearably slow to run but it allowed me to do the rhythmic experiments I wanted to do at the time, mostly high-order tuplety things and pieces with multiple simultaneous tempi. But I quickly learned (as did so many others) that simple and complex rhythms alike sound quite bad when played perfectly. I found myself very quickly editing all the microrhythms by hand, moving events a few milliseconds this way and that. I did gain some intuition about what seemed to work, and I did see that such small changes - as really any performer knows either intuitively or consciously - make all the difference in the world. After a while I tired of all this detail work and went back to having people play the music, which is easier but I suppose a way of avoiding a commitment or responsibility.

But my new opera, Mordake, is an electronic piece and thus unconstrained by some of the limitations I have lived with for a while. Also, I've gained a certain self-awareness over my personal half century that will hopefully allow me to forage for myself without worrying about phantoms peering over my shoulder. So I look forward to Nancarrow-like improbabilities, irrational tempi, manipulations and retrogrades and unplayable parts of all kinds. Amen.

2 comments:

Bill said...

Spendid! I was captivated by the concept of concurrent (divergent) time signatures in my college days('70s) and have continued to have dalliances with such techniques, mostly leveraging MIDI sequencing (bless you, Dave Oppenheim!) but finding the same thing your s-100 experiments revelaed - that machine perfection in applying these juxtapositions doesn't make for musicality and micro-tweaks are a necessity. Not that that isn't true in rhythmic terms in general (strict quantization is a curse). Your [rhetorical, perhaps] question about classically-trained musicians seeming incapable of grokking the intent of the composer caused me to flash back to a piece I wrote while still in school that elicited an awful reaction from both the conductor and the [student] players The last movement of the piece took the themes of the previous 3 sections (in 7/8, 11/8 and *argh* 37/8 respectively) and layered them. Despite what would seem an easy touchstone (the common denominator), it was more challenging than anyone (other than me) cared to content with.
Does your recent work in this area involve non-coincident tempi? That aspect I haven't explored to the extent I'd like to, as it seems to be just as difficult to achieve with sequencing as with human performers - a reference to some other timestamp (SMPTE?) need be employed, afaik.

bt

Erling Wold said...

No, I haven't worked in different simultaneous tempi with real humans playing for a long long time. But I've been toying with it again recently and I have a probably unfulfillable desire to write something for multiple orchestras and of course there would be multiple conductors and multiple tempi. I happened across John Greschak's site when googling around and he discusses quite a number of 'polytempo' works that do all this and more. Check it out.

Related Posts with Thumbnails