New York-based alto saxophonist and composer Steve Coleman is not only creating a musical language for the '90s, he's literally programming the computer musician of the future.

"IT'S A COMPLETELY DIFFERENT APPROACH - and I know it's different because a lot of musicians have trouble with it when they try to play in my band or to play my type of music."

Speaker Steve Coleman is in the UK with four musicians who obviously don't have any trouble with his approach to making music: the current line-up of his band Five Elements. It's November of last year, and Coleman has flown in from New York with keyboard player James Weidman, guitarist David Gilmore, bassist Reggie Washington and drummer Tommy Campbell for three dates to coincide with the release of their fifth album, Rhythm People (The Resurrection of Creative Black Civilisation). Hopefully it won't be too long before they return: their performance at London's Hackney Empire is a revelation, conveying a rawness and cataclysmic energy which lifts the music beyond even the excitement level of the album. This is what live performance is about: musicians stretching themselves and the music, playing around with it rather than just playing it.

Two days earlier at Coleman's hotel, the 34-year-old saxophonist is keen to get one thing straight: "A lot of people have said that what we're playing is a kind of fusion music, which I take as an insult. For us, it's much more closely related to African music in the sense that everybody has a space, a certain place where his part fits with somebody else's. I don't care if you don't know where one is, all you have to do is know the relationship of your part to this guy and to that guy and to that guy, and of their parts to one another, know these different relationships and know how to explore and manipulate them. A large percentage of the music is improvisation, but I try to structure it so that there are different levels. We call them 'rates of change', which is a concept I got from African music."

Coleman uses our interview as an analogy to explain 'rates of change'. It goes like this: when he's answering a question, he's improvising at a fast rate of change, whereas I'm improvising at a slow rate of change because I'm listening to him and perhaps making comments and interjections. When I put a question to him and he's listening to me, I'm improvising at a fast rate of change and he's improvising at a slow rate of change. Simple - or is it? How do Coleman's musicians develop the 'knowing' which he referred to earlier?

"To begin with, I give everybody parts, and if everybody plays exactly what I wrote then they'll get the style of the piece without me explaining it. If they play their parts and it all hooks up in a certain way, the different melodies and rhythms are contrapuntal in such a way that they create the total sound that I want, and if we keep playing it and keep playing it then they'll hear that sound. Once they hear that sound, if they're the right kind of musicians, they'll be able to understand how they can improvise and still keep those relationships.

"With a lot of musicians, if they jam they have to play some tune from the '40s. What I'm trying to develop is a common language for today, using the elements of today, based on a balance of improvisation and structure - highly-structured improvisations, in fact, dealing with a lot of structure as opposed to what the free jazz guys in the '60s were doing."

The two most extreme forms of music in the 20th century, free jazz and total serialism, both succeeded in atomising the elements of music, one through an excess of freedom in improvisation, the other through an excess of rules in composition. During the past 20 years, jazz has lost its radicalism, as jazz musicians first of all fled into the populism of fusion, jazz-funk and the music's own past, and then, with the exception of all but the most adventurous musicians, ignored the radical and exciting developments which took place in popular music during the '80s. Perhaps because the '80s were above all the decade of computers, sequencers, drum machines and the 'non-musician', jazz musicians, with their emphasis on technique and 'real' musicianship, seemingly didn't feel that hip hop, house and techno with their explorations of rhythm, sound, texture, space and noise had anything to do with them. Yet no music was more radical in the '80s than Public Enemy's.

Through his record and production company, C & M Music Productions, Coleman has been producing hip hop alongside his own music. His inclination is not only to deal with the music but also to try stretching the form of it, bringing his own ideas on rhythm and structure to bear. He's quick to point out that in the past the development of jazz was not isolated from the popular music around it, and puts himself in that tradition.

"Soul music, funk and rap are to me what blues and rhythm 'n' blues were to Charlie Parker", he maintains.

"I base the feeling of my music on that. Then I grab the formal elements from music all over the world, from classical music just like Bird and all those guys did, but also from pop music and African music, Bulgarian music, whatever, and then I build an improvisational language and a sophisticated kind of structure based on that. In that sense I feel I'm doing exactly the same thing as Charlie Parker did, though whether I'm doing it as good as he did is another matter - that's something only time can judge."

Coleman grew up listening to soul, funk and rhythm 'n' blues in his native Chicago. When he took up the alto sax at the age of 14 it was to model himself on James Brown's alto saxman Maceo Parker. He gained his initial performing experience playing James Brown songs with a local funk band, and used to copy Maceo's recorded solos note-for-note.

It was his father, a confirmed Charlie Parker fan, who made sure that he heard jazz as well as funk and soul. When Coleman went on to study music at university and tried to join the jazz band there, he was told that he would have to learn to improvise first. This he did by learning Charlie Parker solos off record - a daunting task - and by sitting in on jam sessions in Chicago with local jazz musicians.

In 1978 Coleman headed for The Big Apple, where he played on the streets and lodged at the YMCA until a place in the Thad Jones/Mel Lewis big band took him to Europe. Subsequent work in the big bands of Slide Hampton, Cecil Taylor and Sam Rivers gained him more playing experience but didn't give him an outlet for his own musical ideas, so he founded the M-BASE collective with a group of like-minded musicians based in Brooklyn.

"The original people - me, Cassandra Wilson, Greg Osby and Geri Allen - grew up in the same way, in the same period, listening to the same kind of music", he explains. "We didn't have exactly the same goals, but we wanted to do similar things in creative music. What was important for us was the common base that we had."

M-BASE apparently stands for Macro-Basic Array of Structured Extemporisations, which Coleman not surprisingly admits is "a heady thing, but it best describes what we're trying to achieve."

Along with the Five Elements albums, Coleman has also written two film scores, produced albums by such artists as Geri Allen, Cassandra Wilson, Strata Institute and Steve Williamson, and recorded with the likes of Abbey Lincoln, Branford Marsalis and Dave Holland, appearing on three albums by Holland - including the recently-released Extensions on ECM Records.

LISTENING TO THE MUSIC on Rhythm People, it's clear that Coleman's music is based around a contrapuntal interplay of melodic and rhythmic motifs.

"I don't call them motifs, I call them cells, but yeah, I'll take a cell and I'll develop it, whether it's a rhythmic cell or a melodic cell or both", Coleman says. "I'm conscious of the interplay between different melodies, and as a result of that interplay a sound develops. You could say that that sound is harmony, but I'm not thinking explicitly of harmony, I'm really thinking melodically and rhythmically. So I have a sort of pre-harmony way of thinking."

Coleman's music has a strong rhythmic base coming from the drums, but the constantly shifting bass and snare hits have little to do with the constant reinforcement of the beat common in the unrelenting 4/4 of Western popular music.

"I don't believe in 4/4, and in that sense I don't believe in metre", he says. "Four-four isn't the natural rhythm of things, it's a very contrived thing. I'm talking to you in a certain rhythm, in a certain phraseology, but what time am I talking to you in? Is it 4/4, 3/4? In Europe they had this march thing going, and that's what became dominant. For me, 4/4 time is part of the European influence on music.

"Rhythm was as developed in Africa as harmony was in Europe, but many people haven't realised that because Western culture dominates the world now, and as a result Western values are dominating the world also. But that's logical. Any culture that conquers another culture imposes its own values on that culture, just like Rome did with Greece. As a result, the things that were dominant in the other culture become less significant in the new culture. In Western culture, things like 4/4 time and harmony are really dominant. For me, it's rhythm first, then melody, then harmony if it exists at all, but the way it's taught in school is exactly the reverse. To tell you the truth, I don't even believe in harmony."

Coleman takes issue with the jazz teaching orthodoxy which says that the main contribution of Dizzy Gillespie and Charlie Parker was harmony.

"If you really listen to the music, it's a melodic and a rhythmic music, but the way that it's analysed in school is the same way that they analyse Western classical music. They've taken those methods and applied them to music schools like Berklee, and that's how a lot of these young guys are thinking now.

"I'm not saying that guys like Charlie Parker knew nothing about chords, but what is harmony? At one point polyphony got so complicated in Europe that people decided to come up with names and rules and generalisations for some of the instances of what was happening. So they had names like tonic and dominant for what was really in the beginning complicated polyphonic moves. They just codified them into a harmonic language.

"Harmony as a theory was really only developed in Europe. All the music in the rest of the world is basically contrapuntal if it has more than one line, though of course there aren't the same rules as there were in the music of Europe during the Renaissance. Africans aren't concerned with the same rules that you would apply if you were analysing a Bach piece - they have their own rules, their own logic to their music."

TO STEVE COLEMAN THE MUSICIAN, composer and producer can be added Steve Coleman the computer programmer. Coleman has enthusiastically embraced computers ever since a friend suggested back in '85 that he buy a computer and a sequencing package instead of a cassette multitracker. After investigating what was available, he opted for a Commodore 64 and Dr T's Keyboard Controlled Sequencer software, and also joined a Commodore users' club to discover more about programming. He was interested in the golden-mean ratio 1:1.618 - otherwise known as the golden section - which has been regarded as an aesthetically pleasing proportion since the time of the ancient Greeks, who derived the relative proportions of their architecture from it.

"I figured if you could use the golden section with numbers then you could use it with notes, because everything's numbers to the computer anyway", Coleman says.

Before long he was writing a program in 6502 assembler on the Commodore 64 which was able to expand or contract a melody according to golden-section proportions. From there he developed the program to improvise a melody in the golden-section style and another style which he'd developed himself and called Symmetry.

"The Symmetry style came from watching the flight patterns of certain animals like bees and hummingbirds, which had this jagged flight pattern to them", he elaborates. "I came up with this musical style which imitated these jagged shapes through music. I was originally an artist, and I still see music visually.

"Basically, Symmetry is a technique that, instead of dealing with tonal areas, deals with what I call focuses, areas that you can expand into and contract out of evenly on both sides - hence Symmetry. I first developed this about ten years back, way before I started using computers, but I thought I'd try implementing it on the computer to see what would come out. In fact, the computer started doing things that I didn't understand at first. I printed out the figures and it took me about an hour of analysing them before I could follow the logic it was using. This is hard to explain without going into an explanation of the theory, but basically the computer showed me how to nest focuses. It actually taught me how to realise something that I made up in the first place and programmed into it - it's like being influenced by a guy who doesn't exist!"

"Rhythm was as developed in Africa as harmony was in Europe, but many people haven't realised that because Western culture dominates the world now."

Coleman's next programming step was to get the computer to mix his golden section and Symmetry styles according to a user-specified percentage; to get the program to improvise in real time. He had to progress to an Atari ST and begin programming in FORTH and 68000 assembler. From there he developed the program to improvise over any series of chords played into it from a MIDI keyboard.

"It could play complicated songs like 'Giant Steps'. It would be playing the changes just like a musician would, but it wouldn't be playing any cliches because it was playing in these two styles, Golden Section and Symmetry, or a mixture of both.

"The point I'm at now is that I've made this chord thing real time. If you start playing chords on the keyboard, the program will follow you instantly, whereas before it had to go away and work out what to do. You can play as many chords as you want, and the chords can be as complicated as you want, the program will follow you, you can't outrun it. I've tried to stump it, 'cos jazz musicians will say 'let's see if it can follow this'.

"What it's really doing is it's relating to what you're doing in terms of sound, not chords. It improvises more the way I improvise - I've just used myself as the model. I don't use a scale-over-chord approach like a lot of musicians do, I just look at it in terms of sounds. By a sound I mean a pitch or a group of pitches. I play in relation to which sounds have the least tension or the most tension against certain other sounds."

Coleman's program now also contains software which improvises rhythms in various styles and time signatures, constantly changing the rhythm in the way that a drummer would:

"It's interesting for me to say 'play a funk beat in what most people think of as 4/4, now play the same type of thing in five or in 12/8 or in seven... The average musician can play a funk beat in 4/4, but if you ask him to do it in 12/8 he can't do it. For me, all these times are the same, so you should be able to do that.

"I did the program as a demonstration. I've always wanted a drum machine that could improvise, and I know from programming that it's not that hard, but these companies like Roland and Yamaha probably think 'where's the market in that?'. They could be a lot hipper, but they aren't because they tend to make things that appeal to the average musician. I think all drum machines should improvise and have algorithmic thinking-type things in them. Roland are just starting to get there with human feel, but..."

Coleman acknowledges the pioneering work of fellow jazz musician and computer programer George Lewis in the field of intelligent interactive improvising software, and says that Lewis showed him his specific technique for simulating intelligence within a computer using a random number generator.

"I took that basic idea and ran with it", he says. "It opened up a whole new world for me in terms of the computer-seeming-like-it-thinks thing."

According to Coleman, one interesting consequence of having a computer which seems like it thinks (which thinks that it thinks?) is that it has the capacity to surprise even the person who programmed it.

"Although I programmed certain rhythmic tendencies into the program, once in a while it'll come up with rhythms that I wouldn't have thought of or that drummers wouldn't play. Those things have found their way into my albums."

Coleman uses his program to provide raw material for him to work on. He has a version of the software which can be loaded up into Dr T's Multi Program Environment, where it can run alongside KCS and write its improvisations directly into KCS's tracks.

Today he uses two sequencing packages: KCS and C-Lab's Notator.

"Between them I think that's all you need", he opines. "When I'm doing rap things, SMPTE-intensive things in studios, I use Notator because it's faster. I don't like the way Notator is controlled with the key, though. I ended up getting two Notators just because I didn't want to switch one key back and forth between my ST and my STacy.

"Notator's so easy to use that a lot of people feel it's the best way to go, but for my own music, I use KCS most of the time. I write in a lot of weird forms with all kinds of different time relationships, and the Open mode in KCS gives me the structural flexibility I need because it allows me to play sequences from the keyboard and to have sequences call other sequences. You have to be into the number thing to get the most out of it, 'cos the program's kind of computer-oriented, but numbers never bothered me."

Coleman sees his software as playing a different role to that of a sequencer.

"A sequencer is just a straight tool like a hammer and I'm building a house with it", he says, "whereas my program is more like a creative tool, I can get ideas from it, I can put other ideas in and get feedback from it, things like that. It gives me different ways of looking at things. I can leave it running for an hour and it will just keep coming up with all kinds of ideas which I can print out, write out, whatever. It's almost like you have to get away from it after a while because it comes up with so many ideas that you can't absorb them all.

"I use my program at home to work on ideas, but I'm still not at the stage where I'm bringing computers on stage and playing with them, because I'm still playing with people and I still like that kind of thing."

However, all that might change if one particular project comes to fruition. Coleman is working with Martin Hurni, inventor of the Synthophone MIDI sax controller, on implementing his software on a chip inside the body of the Synthophone.

"I have these dreams about this future kind of music", he says, "and for me an instrument like the Synthophone is moving in that direction, with what we have in mind. I'm not interested in playing an instrument that's just going to make a synthesiser track my sound. That's boring to me. I want it to have some kind of intelligence. I'm interested in playing something and having an instrument play something else as a result of what I've played, something that's coherent, structural and musical, not just random notes. If we can get it to work then I'll start playing the Synthophone on stage."

Coleman sees the musician of the future as "someone more like a George Lewis type of guy, a guy who's very versed in music, very versed in technology and very versed in computer programming."

There again, there's a limit to how much time any one person has. Most of Coleman's time is taken up with playing, composing and producing, and as a consequence he feels that he could develop his software more quickly by bringing someone else in to work on the details of the programming.

"So if there's anybody out there who's interested, maybe they could contact me through the magazine", he suggests. "I realise there are people out there who are only interested for their own gains, but my ideas are... I feel they're unique because they're musical ideas first. I have a lot of musical ideas about relationships and I can translate them into computer language. I know a lot of MIT-type guys who think in a very computer fashion, in a very algorithmic way about how to change melodies and things like that, but it's not musical first. I program from a musician's point of view: I don't think about the processes first, I think about what I want to do as a person, and then I think about ways of approaching that with the computer, and ways that the computer could do it differently."

If life mirrors art, there can be no better instance than Steve Coleman. His life seems to be every bit as multi-layered and multi directional as his music - and to develop every bit as quickly.

"I've come pretty far in the past five years", he reflects, "and I hope in five years' time I'll be equally as far on from where I am now. I don't want to be one of those musicians that just stays on the same spot."

Somehow this writer can't see that happening.