In 1977 I moved to San Francisco to study photography at the San Francisco Art Institute. After a couple of years there, I was drawn into the rock and roll/punk rock world and quit school to play in bands. That arena soon proved to be too boring so I got out. At that point I met a guy named Mark Pauline who had been doing "machine performances" with different machines he had built himself under the name of "Survival Research Laboratories" (S.R.L). I found this quite intriguing as I'd had an avid interest in machines since a young age. When I was a kid I worked on outboard motors for boats a lot and, by the time I had turned 12, had my own car. So I began to work with "Survival Reseach Laboratories" under the pretense of producing soundtracks for the shows.

After working on several soundtracks, I began to build my own robots. The first one was titled "The Jumping Machine". It stood about 6 feet (ca. 190cm) tall, was pow-ered with a gasoline motor and could jump about 2 or 3 feet (ca. 60cm) in the air. It had a sheet metal skull and a constantly masti-cating jaw.

Over the next six years I continued to build robots of in-creasing complexity, learned to infuse them with some personality, experimented with pyrotechnics, and continued to produce the soundtracks for the S.R.L. performances.

My first attempts at soundtracks were simple and crude, often consisting of just two or three cassette tapes. But as the performances grew larger and more heavily scripted, I wanted the soundtracks to have a stronger relationship to the action taking place during the performance. As a concept this was not difficult as it only required spending a bit more time working and thinking about the upcoming performance scenario and what the "voices" of the machines should be. What I wanted to get happening during the performance, as I saw it, was a layering of reality and suggestion. The "reality" being the sounds I used as the "machine voices" and their actions and the "suggestions" coming by way of dialogue I grabbed from "B"-movies or religious TV shows and mixed in such a way that people who never intended to speak to each other would have emotional exchanges, and somehow this would imply what the machines had on their minds as they carried out their actions.

Having the sounds aligned with action in a performance was difficult, since, even though there was a script to follow, timing was variable and often spontaneous changes took place because devices functioned differently than expected. The way I resolved this problem was by putting various elements of the soundtrack on individual tape loops and have 15 to 20 of them running at once into a mix board so that appropriate sounds could be pulled up at any time.

I worked with S.R.L. from 1980 untill 1988 and during those years machine construction and soundtrack production were separate activities for me. But I had a growing desire to have all the sounds produced by the machines themselves. I also wanted to discard the sounds that had become ever-present at the shows, namely explosions and loud gasoline motors, with which I had become tired. Realising that the only way to achieve this was through independence, I left S.R.L. in 1988.

- to build machines that were "playable" in that each would have a variety of timbres or rhythms that they could produce,

- that the machines be remote-controlled,

- that performances would consist only of sounds produced by the machines themselves - no tapes, samples or any other auxiliary sound would be used and

- the machines' movements would be interfaced through a computer to allow for a finer degree of control and to allow them to appear as autonomous entities.

So now the situation is such that even though the machines' movements can be pre-programmed, I can take control of them at any point. All the sound is produced live by machines without humans obviously screwing with them. It is very important to me that any time someone hears a sound at a performance they can see it being produced - sometimes in a manner which suggests a lack of control or even that something might be about to break.

The types of machine instruments I've built to date tend to be more percussive than melodic. One is the "Oscillating Rings". This is a steel table ca. 240cm x 240cm x 6mm on top of which are 4 rings of 60mm steel tube bent into ca. 1m diameter circles. Each ring is held at an angle to the table by a bearing and driven in a rolling motion by a motorised counter-weight. The motors are variable speed (10-150 rpm) and independently controllable. For each ring there are 4 steel wires that can be moved into it's path of travel to create rhythms as the ring passes over them. The piece weighs about 550 kg.

Another piece is the "Disc/Cable Mechanism" or "Thing on a String". This is an example of something I built for for a specific building, namely: the University Art Museum/Pacific Rim Archive at the University of Berkeley. It consists of a 10mm x 16 meter steel cable fixed to the wall at one end and attached to a motorised rotating arm which can swing in a 60cm circle at the other. The cable passes through a 120cm diameter x 4mm aluminium disk with a 60cm diameter hole cut out of the centre. The disk hangs about halfway along the cable and, as the cable rotates, the disk swings in a large circle on the cable. The rotation of the cable can be suddenly stopped or reversed which causes the disk to bounce quite a bit and make some good sound. Aluminium is shaved off the disk during this proceedure and falls like light snow.