Q1: On P53 CD you set improvised music into 25 tracks, to make them easier to move around or to isolate, study, re-sequence, shuffle,cycle, or skip. What do you think about the work of dividing the continuous piece into tracks when you recorded it?

If you listen to the CD, the musical performance is still continuous so, as far as the ear is concerned, the existence of 25 discrete track points makes no difference. After all, music is sound and track marks are just visual information. Putting track marks in is no different than printing a novel in the form of a book and numbering the pages instead of printing the whole thing on one long scroll. In that respect, I can't see any disadvantages in marking tracks. Are there any advantages? Firstly, there's convenience: tracks are like map references, they can help you find your way around and they are useful if you want to listen to a certain part only, or as a study aid. They also give the listener more access to creative listening. Secondly, it has to be noted that the greater continuity that conditions a live performance - that of shared place and time and of contiguous and co-temporal human presence - is already erased in a recording, as is information from all the senses but one. So a recording already embodies a violent break with the continuity of a real event.

Q2: Mr. Yoshihide Otomo said in an interview, "Mr. Cutler reconsiders what he did during the performance after it has been staged." Do you usually analyze performance afterwards so as to set the tracks on the CDs?

No, I don't so much analyse as re-listen. I want a record to make sense as a record, which is to say as a pure listening experience. So although P53 may have improvised at its concert in Frankfurt, as soon as I listen to the recording, what I am listening to is already no longer an improvisation: firstly because it already exists, secondly because as soon I have listened to it more than once, it begins to become a composition for me. And the more I listen, and the more familiar I become with what is going to happen next, the more of a composition it becomes. This will be true for every listener. So my concern with a recording is never as an improviser and always as a composer. Each decision I have to make - is this recording worth working with, how coherent is its structure, does its dramaturgy work, how might it be improved? - is concerned only with the recording as a (potentially) coherent listening object, in fact one that will have to bear repeated listening - in other words, a composition. As a composer, then, raw documentation only interests me in so far as it is raw material. Of the event at which it originated, nothing remains but an enigmatic stream of disembodied sound. Not just any disembodied sound, of course, a sound full of meaning and musical logic, but not the logic of the concert. At the concert the sound was situated in a context - that context is gone; at the concert the sound was only one of many sources of sensation, now only one remains. And what remains can not appeal to what is lost. So, for me a recording has to work on its own terms independent of any temporal or geographical source. It is either a coherent sound event or it is not. And that is why, as far as I am concerned, documentary verity doesn't justify a record release. Documentary verity is for historical study, or showing off ("we really did this") or conceptual art - all cases where its object is other than to be music.

I could invoke a principle here: that an artist should always respect the medium he or she is working in, and for me the medium of recording -because of its infinite repeatability - is one of composition. A released record is more like a film than a play, more like an object than an event, because it refuses to go away. For this reason recordings demand a special kind of attention. A record listener hears a sound object where a concert-goer experiences an event; a record compiler, when working from a pre-existing recording, becomes a consumer of the sound, where a performer is always its producer. In other words, faced with a recording, both producer and consumer are equal - as consumers. Now, in the case where I am considering my own recording for possible release, I have to move to phase two, which is where I take the opportunity to recompose what I hear using the existing recording as my raw material. I could approach this work with the intention of retaining and enhancing what I perceive as being 'already there', in other words it could be a work of interpretation; or I could decide to make something radically different out of it. Whichever I choose, the reason I do it at all is not because I want to originate new material but because I am claiming composer's rights over material that already exists. The point I want to make is that anybody could take this recomposing approach to any existing recording, and, in this sense at least, all consumers are potentially producers - just as all original producers of a musical performance are reduced to being consumers when they are confronted with a recording of their own performance. Because a recording is always already there.

So, why use live improvisation as a source for a record? One reason is that music produced in real time for a real audience often has complexities, subtleties and structural qualities that could never be reached through composition or interpretation. It may well be - in my case it usually is - that what I am interested in is to find and enhance the musical logic that was produced, largely unconsciously, in the deeply charged and focused environment of a real-time public performance. And in this case, to recognise a potential record in a recording of an improvisation, and then successfully to shape it into an actual record - like Michaelangelo seeing a sculpture in a block of stone and cutting it out - is a compositional skill, and should be approached as such. A poor performance might still make a good record; equally a great performance might make a bad one. So, never blame the players for a bad improvised recording (unless they compiled and approved it); it can only be a bad composition.

Q3: On the contrary, there is no such division on "Fred Frith / Chris Cutler Live 1,2", mixture of his performances in the 70s and two of yours' in the 90s. At what point do you make a distinction whether setting tracks or not.

I think I may have put index marks in earlier CDs, if not track marks. But if the piece is short I would probably not break it up. Also, on earlier records I just did not think about this question, so it was not a conscious decision not to add marks. Now I would break up any long piece, just for listener convenience.

Q4: When we re-arrange the order of the tracks on CD, the piece may be substantially changed from the live performance. Re-arranging the order may cut off the flow of the sound and combine previously unconnected sequences. What do you think about the relationship between the live performance and the recordings on the CD?

Re-arranging the tracks will destroy the structural integrity of a linear, unfolding piece, it will dislocate a given structural logic. This might not suit the original composer but, as I have already said, since a recording is not only the trace of an event but also and always raw material too, then for a listener, the ability to experiment with random or planned dislocation might well have a value - for instance in that Cagean, Zen sense of unhooking reflex and habit in order to arrive at something unfamiliar and surprising. Once you have listened to the p53 CD in its intended order a few times, if you then play it in shuffle mode it's extraordinary how unrecogniseable much of it becomes. Even sections you know well sound different when they emerge unexpectedly from the 'wrong' place and then fail to develop in the manner to which you are accustomed. When each listening is different in this way, the listener is forced to continue to listen and not allow the brain to switch into habit mode. It makes you hear more, and hear differently. So, in this sense at least the shuffle function could be seen as a valuable tool. I say tool because I am not sure whether experiments like this have a musical value or not. They certainly have an educational and philosophical value; and raise highly provoking questions about structure, intention, expectation, chance, coincidence, the psychology of listening and the objectivity, or subjectivity, of order.

Q5: In "June 12 1998 at the edge of chaos", you shifted the electronic instruments though the sound did not change much. How do you see the connection between gestures and sounds? The only thing we can do with a CD is just to listen to it, even though we can't see what the performance is like.

Exactly. A record is only a sound object. All records are by definition acousmatic - which means that the source of the sound cannot be seen. Since a record only speaks to the ear it inevitably renders most traditional performance values, such as gesture, the shared creation of time, the physical force of sound, universal amplitude and timbre, meaningless. That is why, in the age of recording, these same values have become increasingly important; it is their confrontation with their own absence that has conferred an enhanced semantic and sensational force upon them . What else could live music offer in the face of perfect and perfectable recordings?

So, in a performance I care a lot about the relation between gesture and sound - and that is why I have a problem watching motionless people hunched over laptops while World War Three pours out of a PA system; I just can't help wondering if the sound is coming from them or whether I am just listening to a CD. So, while in the case of traditional instruments there is a clear and manifest relation between the information that reaches the eye and the information that reaches the ear, with laptops and samplers - even when they are played 'live' - the experience is still an essentially acousmatic . And for me, that is a problem. Certainly, I evolved my electrified kit with a view to maintaining a clear relationship between what I have physically to do in order to produce a sound, what the public sees and what all of us together hear.

Q6: I think the important point of your play is the process of creating sounds. Do you see CDs as supplement containing this process? There are a huge numbers of combinations of track order, and still I think CDs are not completed product. Or, would you say it is a part of extension of your play?

As a composer I want you to listen to my work as I intend you to listen to it and in a concert, this is almost possible, since we are all in the same space at the same time, listening to the same thing. But as a listener - and as the producer of a sound object - I am also happy that the medium makes it possible to change the volume, equalisation or order of tracks. This doesn't make my work disappear, but it does recognise the possibility of other ways of listening to it. So what is lost in the shared construction of an event, is at least partially compensated for in a recording by relinquishment of absolute authorial control over the material.

I would say that a performance asks - at the listening level, I do not speak of functional musics here - for an initial atmosphere of hospitality, even formality. This is a human relationship, artists are presenting their work and in general it should be received with courtesy. But there is no a human relationship to be had with a record, it is always a meeting between a person and a thing. The artist is not present to be insulted, no matter how the listener behaves. And, the only listener the artist ever related to when the record was being put together was him- or her-self . So, how to approach a record? I would still counsel hospitality, but the fact is that the record belongs to the listener and that is the essence of the relationship. The artist has made an object - a commodity in fact - and he or she must be prepared for it to be treated as such.

Playing a record creates no new material, it merely enables an interpretation or a reinterpretation - or at the very least an experience - of a thing that already exists. In this respect listening to a record is like reading a book or looking at a painting, it is a communication conditioned by absence. A performance on the other hand always embodies some creation and is always conditioned by the advantages and disadvantages of presence. Both media share the dislocations of interpretation - that is a basic constant of all communication - but while at a concert interpretation, for the listener at least, can only ever be experiential, when applied to a recording it may just as easily be practical - for instance shuffling tracks, equalising the sound or adjusting the volume, and these are only the easy and commonplace options.The next step is the one John Oswald took in his Plunderphonic work - where he uses other people's already finished recordings as the raw material for his own manipulations and re-composing. Although his extensive electronic and mechanical manipulations extend a light-year beyond the small adjustments of tone, volume and track order that most listeners apply, his work nonetheless remains on a continuum with such adjustments. They can all be thought of as variant ways of hearing. By releasing the product of his own re-workings on a record, John is both sharing his way of listening with us and encouraging us to listen - which in this instance means to process and change his or any other recordings - in a similarly active way. Because it is possible. I think it is not the original performance that such manipulations extend, but the act of listening.