pierre boulez
page two

Stravinsky warned against the university education. Would you?

No, not against the education, not at all; although I'm not really expert enough to have a judgement here. Only that if you could go outside more, if there were more ties with the city itself, I think it would be for the better. I spent only a semester teaching at Harvard, and I looked at the relationship between the Boston Symphony Orchestra and Harvard—this was thirty years ago—and there was absolutely no connection. I find this detrimental for both sides, because you have a very high intellectual level on one side, a very high professional on the other side, and they would benefit from closer contact. 

Has your view on electronic music changed since you began IRCAM? Has the course of electronic music been surprising?

Well, What was surprising, what was quite unforeseen when I began IRCAM… I began the plans for IRCAM in 1969 or 1970, quite a long time ago. I had contacts in New York, and I must say that I had contacts with Max Matthews, who was at Bell Laboratories at this time, in New York. Although he never ordered me to do anything, through our conversations he made me aware that having a room for computers was very important. That's all that I knew, both intuitively and through speaking with him. But I was careful, because at the beginning one can't be sure; I was careful not to give everything to the computer. But progressively, and much more quickly than we had ever thought, the computer invaded everything, from the analysis to the synthesis of sound to the manipulation of instruments—everything. It's a tool which is very general and which can be used in very different ways. The evolution of IRCAM is thus closely tied to the evolution of computer technology. 

The second thing, which was not surprising to me at all, because I had always pushed in this direction… I had had bad experiences myself in the 1950s with electronic technology. If you did something electronic [as a composer], you had something on tape. Then [as a performer] you had to follow the tape; you were absolutely squeezed into coincidence with the tape. It was completely detrimental to the performance. Therefore I pushed the research in IRCAM towards live electronics and live computer systems, so that the computer would be geared towards the concert situation, so that the computer would have an instant response to the performer. 

That was my first push in this direction. The second was the attempt to make the language [of computer programming for composition] more intuitive for the composer. I remember that when I learned, or when I tried to learn, computer music, there were only figures, figures, figures, which don't speak at all to a musician. If you see a figure in Hertz, or in the number of decibels, or if you have to wait half an hour before you have a sound, you are discouraged completely. So for me, what was important was that you could at least make a sketch very rapidly. I wanted the ability to sketch with the sound first, to have sound instantly, even if you refine it later. Also I wanted the ability to use graphics as an instant notation, even if approximate. The musician's imagination is stimulated only by a language which speaks clearly to the intuition of a composer. So those were the two things that I felt—and still feel—responsible for. 

Two recent large pieces, Répons and …explosante-fixe…, make use of electronics. How are they related? 

Originally, I did Répons with the technology of the early eighties, and I abandoned the piece because I wanted a newer technology, especially one that uses the new MIDI pianos which give so much more data than was possible in 1981 and 1984. Therefore, I left that; I will write, later, another part of Répons which will use some of this newer technology. 

The central role of the flute in …explosante-fixe… is something that comes from a long time ago. When I was trying to do …explosante-fixe… first in 1972, the technology was so primitive, you can't imagine. You had still the connections with wires, and so on. It was clumsy and unreliable. I cannot explain the frustration we had. In the period that followed the first version of the piece, I thought always of that. Also, I was working with a flutist, who died in between the two versions of the piece, unfortunately, at a very young age. He was also very interested in connecting instruments to computers. He invented this kind of contact with the flute, in which the flute is registered with the computer immediately. The note you play is registered not only with the keys but acoustically, so you have the two aspects electronically controlled. Progressively, I pushed in the direction of a “score-follower:” this is the computer's capability to follow the score just as the flute plays it. With that, you can then trigger whatever you want: you can not only modify the sound of the flute—which was still interesting to me—but you can also link the flute and the instrumental score to a third part, an “artificial” score. Currently, I'm working further in this direction. I want the computer to read the data of the flutist's performance to simultaneously modify the artificial score; for instance, if the flutist decides to play very slowly, then the artificial score will be very slow. The idea is to achieve an interaction between the player and the machine. 

…explosante-fixe… has a complicated family tree, doesn't it? I'm trying to remember all the pieces that came from it: Originel, Rituel, Memoriale… 

Rituel is different. It is based on just one chord, that's all, which I took from the others; so it's just the harmony. But yes, I did …explosante-fixe… first for seven instruments; and then I saw that that was too complicated, and the machinery was so terribly inefficient. But each part was written, so I have now a reservoir of possibilities which I will exploit progressively. I began with the flute. There was a violin part, as well, and I have begun to transcribe that part for a work for violin solo [this work became Anthèmes and was premiered in New York by the Ensemble Sospeso in 1998]. 

A lot of your pieces are interrelated in this way.

Yes.

Absolutely: it's a tree which gives another tree which is another tree. Therefore, the title Dérive. Dérive 1, for instance, is from Répons. Part of the material for Répons I did not use, actually, and that became the material for Dérive 1. And Dérive 2 is from studies I did for the part of Répons which is still not written. Dérive 3 is derived from another piece, Le Visage nuptial. You can never use all the material. But I like these kinds of relationships. As long as the material is not used fully, then I like to have some derivations.