Tod Machover: Technology and the Future of New Music

6. Theremin

FRANK J. OTERI: A piano keyboard is very logically designed… Once you know the system, the analog makes a lot of sense — the seven white keys, the five black keys, the patterns repeat, you’re up the octave, you know what note you’re hitting. On a violin, it’s the distance that you play, and it’s actually more difficult in terms of having intonation. There are no frets; you have to tactilely figure it out and through playing it, you know. And with a trombone, even more so. The extreme of that is probably the theremin, an early electronic instrument which is popular again.


Spectres -- CD cover

RealPlayer  [72 seconds]
RealAudio sound clip
TOD MACHOVER: from Spectres Parisiens (1984)
The Asko Ensemble conducted by Peter Eötvös
from the CD Tod Machover: Spectres
{Bridge BCD 9002; distributed by Koch International}
Order from Amazon.com

TOD MACHOVER: I think the theremin is a really good place to start this discussion, because the theremin is a completely new interface. It allows you to do something completely new, something that couldn’t be done with existing instruments. It allows you to move in space and control sound very precisely. It’s also probably the hardest instrument ever invented. It’s incredible in possibilities. On the other hand, you’re playing something with as much variation as a violin without any physical reference point. You’re just moving your hand in the air, trying to nail perfect intonation, nail amplitude control without having a bow. There’s nothing to push against. There’s no neck to figure out a reference position. I mean, the violin’s hard enough. It’s no accident that there are not that many theremin viruosi around… There’s Clara Rockmore

FRANK J. OTERI: And Lydia Kavina.

TOD MACHOVER: Yeah, not to slight her. I mean, Clara Rockmore was better. It’s just hard to play the theremin; there’s no question about it.

FRANK J. OTERI: There’s Samuel J. Hoffman, who did all the sci-fi and horror movies in Hollywood in the early ’50s.

TOD MACHOVER: Right. …he was pretty amazing, but he could probably edit what he did. It wasn’t live. It’s just a hard instrument. I think the technical answer to your question is that the amazing possibility of this whole new generation of instruments, composing devices, whatever you want to call them, is that, unlike the theremin, in between what you’re doing physically and the sound-producing device, there’s a computer, there’s software, which means that you can define the relationship between what you do and what comes out the other end. What the constraints are and what the definition of the instrument is yours to describe as designer and composer. And the good news about that, and in fact what I tend to do, is, just like with any musical composition, depending on who you’re designing this for — whether it’s for a 5-year-old, for a virtuoso musician, for somebody who’s got five minutes — you can define and design these constraints differently for every situation. Not cavalierly, but as an integral part of the musical work. I consider that one of my jobs as a composer these days is to design the way the interaction works, the way it feels, what it means to learn it, how much is given ahead of time, how much is free. And so the good news is that this power of being able to design exactly how the instrument works is available to the creative artist. The corollary to that is that you have to define it; nothing is free, nothing is predetermined.

FRANK J. OTERI: Unlike a piano, which has this analog, or even a theremin, you know, the closer you are, the farther you’re away, it might be difficult to realize, but you can conceptualize what that means. On the ball, what is C-sharp? How do you know where F- sharp is? How do you know what you’re doing on it? Or are you just making neat sounds? Where do you draw the line? Where does it stop being a toy and become an instrument, I guess is the question.

TOD MACHOVER: Probably there’s some room to build toys also. I mean, toys are fun.

FRANK J. OTERI: Toys are great.

TOD MACHOVER: Not to be flip about it, but I think there is room to build things which don’t do hundreds of different things, but rather does one thing extremely well… You know, something that might make one kind of sound. I mean, maybe we designed a version of this fabric ball that only made vocal sounds and had an incredible variety of ways of controlling vocal timbre or something that could be vocal to electronic. It did that one kind of thing sensitively and well, and you always knew it was going to be a voice and you weren’t going to make your own composition with it, but you were going to have a lot of fun changing the voice and exploring it. I think there’s room for things like that. My own feeling is that there are a couple of things to learn from the way we normally make music and from traditional instruments that are worth keeping in mind when you design a new one. One is that for any instrument or tool to be interesting, you have to be able to predict what it does and know that it’s going to do the same thing if you pick it up again. Another thing is that you must find a way to isolate the essential things to measure in human expression, and then get a total, integrated picture of what the music-maker is doing, rather than a disconnected series of parallel data streams. Take a string instrument. You could analyze it by saying, “My pinkie, when on the bow, has this kind of pressure, and at the same time my thumb is doing this and then each finger is kind of moving in an independent choreography, etc. At some point when you’re learning the violin, you do isolate each finger and the arm, but the reason the violin is so wonderful is it takes these many, many different motions and weights and activities that can all be integrated, are all coordinated. I mean, you’re moving this arm like this. At the point that you play the violin, the last thing you want to do is think about the arm separately. You learn little by little the way it feels to have all of these actions coordinated, and in some magical way — well it’s not so magical, it’s the way we do everything in the world — you think about everything together. You don’t think about the angle of your wrist and the pressure here and the speed there as being separate. You can’t think of them separately.

FRANK J. OTERI: It becomes second nature.

TOD MACHOVER: It’s not so much that it’s second nature. You’re doing probably 20 different things with your right hand, but you don’t think about them as 20 separate actions. You don’t think about each finger as a separate little instrument up here. One of the big problems with a lot of electronic instruments is that the easiest thing to do with electronics is to say, “Okay, I have a switch here, a switch here, a switch here, and so when I push this, I’m going to send information out to my computer, I’m going to have these 20 parameters, these 20 things, which are clear, because they’re physically separate places here. And I’m going to send them all out to my computer, and they’re each going to control some separate bit in the music or in the structure, and I’m going to give this to somebody, and you’re going to learn it by thinking, ‘Okay, this is loudness, and this is the range of my notes, and this is timbre…” And what I propose to you is that that doesn’t work. That doesn’t make a good instrument. The way to make a good instrument is to coordinate and correlate all these separate functions so that they add up to something integral, which is more than the sum of its parts. Of course you have to have separate functions, but just like a violin, I think the trick is to pick a number of things that you want to measure and to learn how to define and measure their interrelationships. There’s always a magic number around 10 or 15 things. If you’re trying to measure 100 different things on an instrument or that somebody’s doing in a performance, I think you’ve picked up too many insignificant things. That’s too many things to measure. And if you’re trying to measure them as being separate and disconnected, you’ve also thought about the problem incorrectly, I think. Because when somebody picks up an instrument, you don’t want it to literally feel like a violin, but you want somebody to forget about the interface and not worry about exactly where their fingers are going. I think the metaphor of touching something, squeezing something, having something which is simpler if you do something simple with it, something which is more complex if you manipulate in a complicated way — this is the more natural and productive way. You want some relationship between the way you feel when you play the instrument physically and the general way that the music is constructed. So even though it has to have separate controls, you want them to be interconnected — interdependent, we call it — in a very sophisticated way. So you want the number of things that you’re measuring to be not more than about 10. You want them to be just the right things. And you want them to be connected in a sophisticated way. And I think once you do that, you come up with something that feels good for a child or for an adult, and then you fine-tune it. Then the question is, as you’ve asked already, how much in the instrument is predetermined: Does it have scales already? Does it have fragments of music that I compose and put in there? Or is it more like a piano, where it has kind of certain possibilities but no particular direction?

FRANK J. OTERI: You give somebody sheet music to play the ball and there are notations for playing the ball. Do you have such things? Are there such things? Would they work? Could they work?

TOD MACHOVER: Yeah, I think notation is an incredible problem. Its hard enough to learn what it feels like to manipulate one of these new instruments. To learn a new graphical system, to look at something related to what you’re doing physically, to remember it, it’s just a can of worms. We haven’t done much on that.

FRANK J. OTERI: Is it even a valid question anymore? Do we need this? What happens when you’re not around and someone wants to perform a piece like this, say, in Marrakesh or somewhere in Russia?

TOD MACHOVER: I think it depends on the definition of the instrument. If you’re making an instrument — for instance, a fabric ball like this — that’s meant to play a fairly specific line of music for, let’s say, a chamber music piece that I’m writing, then you would have to come up with some kind of notation. It wouldn’t be enough to say, “This piece is 20 minutes long. Squeeze this thing any way you want to for 20 minutes, and something nice will happen.” That wouldn’t be precise enough. But that’s not usually the way I think of making music with something like the fabric ball. Generally, the way we’ve been thinking about such a radical instrument as this is not so much writing a score for it as trying to make it exploratory, to make it feel right, so that the instrument itself conveys the right information needed to make interesting music with it. If you think of it like a sound sculpture or a musical space, with a set of possibilities embedded in it that emerge depending on how you manipulate it, then playing it becomes more like conducting or shaping, like directing the computer through an interface like this. For this kind of activity, you don’t necessarily need a score, because everybody’s going to do play it, explore it, in a slightly different way. If you’re writing a piece for it, you clearly would need notation. And actually one of the big challenges for this Toy Symphony project is that we are making these music toys, we are inviting children to work with professionals as equals, and I’ll be writing a piece for this. We’ll also be commissioning young composers to write pieces for these music toys. We’ll be working with kids as well, so I think we’ll need to think about notations and ways of making things that can be reproduced. Over the next two years, we’ll clearly have to think of the notation issue. And that’s a real can of worms.

You might also enjoy…