George Lewis and Voyager

On Friday, March 30 Professor George Lewis presented a lecture entitled “ Living with Creative Machines; Improvisation, Technology, and Interactivity.” Professor Lewis is an accomplished trombonist who has appeared on over eighty recordings. He has been a member of the influential AACM organization since 1971. He is currently the Edwin H. Case professor of music at Columbia University in New York. He is an influential computer music composer that has produced various sophisticated programs for his work with improvisation and interactivity. (For more Bio information go here.)
The talk began with Mr. Lewis describing his initial thesis and interest of study, improvisation. He made the point that improvisation is all around us and that we use it everyday, we just don’t recognize it as improvisation. When we hear the word improvisation we naturally think of music, but he wants to take the concept into a wider discussion; Improvisational studies. He sees the world as being full of networks and matrices of interconnected knowledge, social exchange, and power relationships. The study of improvisation is the study of strategies of navigation through these networks. The interaction we experience while moving through these networks is how the exchange of meaning occurs. Since these networks are becoming all the more ubiquitous, the study of our interaction with these systems is critical.
To study these interactions Dr. Lewis builds computer-based machines. Dr. Lewis’s field is music so it is natural that he would build a musical machine that simulates the conditions of improvisation. His computer program, Voyager, is an impressive piece of programming. It plays with a live human player and interacts with them as a human might. It becomes a member of the band. And just like any member of any band there is no off switch. This is interesting because a person interacting with this software needs to guide the music with his or her own sound output. Sometimes that means taking charge and leading the music, other times it means supporting the programs lead. These two roles need to coexist and exchange in an instant and the program, remarkably, can do this. Since Voyager is treated like a player, the human player needs to deal with it like another human. This is what Professor Lewis is interested in, how a player deals with the process called improvisation. What are the player’s strategies for steering the music towards satisfactory musical output?
Since Improvisation lies outside of verbal communication a definition of it is hard to come by. This is one of the reasons that there is a lingering doubt about the ability for improvisation to be even studied. He said during the lecture that he can’t really tell you what it is but that he knows when it is happening. This intangible quality makes the clear understanding of improvisation nearly impossible to grasp. Perhaps that is it’s greatest strength. A person needs to experience it not just come to some intellectual understanding. The research with machines like voyager can bring a clearer sense to what it is like to flow with the music by identifying what inhibits that flow that inhibition is in us . To realize it is to know yourself better. It really has nothing to do with the machine. So the point that is central to his work is how these improvisational experiences are reflected into the society at large. If Voyager gives some insight into the world then it has done it’s job.
Befor I conclude this entry I just want to express my opinion about these machines. I think his machines make a good representation of improvisation but that’s it Voyager is not a good improvisor. I think that ultimately improvisation is a human activity and just because these machines can mimic human qualities to a high degree does not make them capable of human activity. I don’t want these comments to seem like I am putting down this program, it is a remarkable piece of work. The fact that this is even part of the discussion reflects positively on the success of this program. But it is not an improvisor it is a computer program. It crunches numbers and spits out sound, that’s it. To think it is doing anything more is an illusion. I think it is a mistake to give it a human character or anthropomorphize it. It just processes numbers. This is not something that everyone needs to discuss, there is a certain suspension of disbelief that takes place with these interactions and that makes the experiment interesting and worthwhile. But we are in school to look at all the angles and learn about what makes this stuff tick at a fundemental level. So I just want to say I think for his studies of improvisation on the social or cultural level it plays an experimental role that delivers useful data, but it is a stretch to call it a musical improviser.
Here are some recordings of Voyager performing at the UWM e-reserve website. Sign in and give it a listen.
This lecture was one of the best I have seen thus far. It was thought provoking and the questions it raised have led my thinking into interesting and personally productive directions. It is always a pleasure to listen to smart people talk about things that interest them.