(I hope this is OK, I'm not making an ass out of myself by posting these ramblings am I? I'm writing this little by little in a background window while I do my job all day, and I'm hoping to contribute to an interesting discussion? I guess I should expand this into an essay and try to publish it somewhere? :-)...

"The Rat's Nest" is a piece off my CD which grew out of a philosophical investigation into "appropriation". I was thinking about various things: perhaps Daniel C. Dennett's concept of "the intentional stance" is a good one to keep in mind. I was looking for the layers of complexity involved in the "intentionality" of "works of art" in the Western Romantic or 20th century sense, the "edges of intentionality" maybe, or for what lurks underneath the rock of intentionality, when appropriation and discovery are used as building blocks for subsequent creativity/intentionality?

Consider Marcel Duchamp's "readymade" urinal or his moustache on the Mona Lisa. Graffiti "art" in superimposition over existing objects. Plagiarism, tribute, "Annotation" in the field of literature.

Consider of course Andy Warhol's soup can or his series of Marilyn/JFK/Mao etc. colored silkscreened images from photos...

Consider Photography itself, the "capturing" of a visual scene on film, in the history of visual arts, going back to the "camera obscura", the observer in the box (however, consider James Crary (?) "Techniques of the Observer" (? I probably got the author/title wrong, sorry, it's on the shelf at home :-) where he claims that photography is NOT descended from camera obscura perspective).

PhotoShop nowadays, and other applications, which can radically alter images.... Forgery. It's big business now. (of course we have something equivalent in digital audio workstation editing, subject to the fundamental conceptual differences between the visual realm and the auditory realm, ref. "Auditory Scene Analysis" by Albert Bregman)

The necessity, in digital communications, for "digital signatures", to "prove" identity/authorship or more precisely, "non-reputability". Will we have to stick "digital signatures" in the headers/resource forks of the digital representations of our "artworks" in the future? And then, there will be layers of superimposition, signatures on top of signatures, embedded signatures found in the data fork itself ....

Then there's the whole question of "animal art", is it "real art", is that factor of "intentionality" present, (or even necessary)? I was just reading the other day, there are apparently elegant coffee table art books of the paintings and drawings done by elephants (they hold the pen or brush in their trunk, no shit! even wild elephants apparently make "artistic" scratches in the ground, unprompted by humans), and art critics have judged the elephants' works to be very sensitive and talented.

Musically, a long ways since the invention of the wax gramophone cylinder used by Pierre Schaeffer, and then the wire tape recorder used by Cologne Radio Eimler/Stockhausen et al, obviously we have blatant "sampling" where a pop group makes a new song based on a recording of an old one (help me out here, I forget the name of the song/group that was based on Rick James' "Superfreak"?).

Or "quotations" appearing in musical works, Charles Ives having achieved impressive results, with of course Peter Schickele's PDQ Bach stuff taking it to outrageous extremes. Larry Austin's recent "completion" of Ives' "Universe Symphony" which itself contains quotations... Or more subtly and more at the heart of our Western Romantic art music tradition (ref. Lydia Goehr's "The Imaginary Museum of Musical Works"), works based on themes taken from folk music, Bartok and Beethoven (the "peasants' dance" in the 6th Symph. Pastorale) come to mind.

I was looking for things that would astound me, surprises, things I could "discover", "use", "borrow", "appropriate", ... I wanted to make a finished piece of music that was brilliant, and which was undoubtedly "mine", but which I could not claim to have created every bit of it "from the bottom up".... so I would almost have to list "anonymous co-composers" or something.

Again, I wanted to do something wonderful and rewarding for the listeners, but still I wanted to humble myself by saying that I am not a "composer", that this could not be a "composition" by any reasonable definition of the word..... and by being silly in this way, I wanted to pay homage, by contrast, to those of you who I respect so much, who really "Compose", by first hearing a sound in your head, then going to work at extracting it out onto paper through notation etc. and eventually through performance to sound.

I know that there have been precedents -- there was an album or two of music made by transcribing frequencies observed in the Earth's magnetosphere (?), I think one of them was done recently by an Italian woman who is an astronomer? Also didn't Charles Dodge do something like this?

And then there was Larry Austin's piece which used the fractal description of a coastline. And of course Cage did his pieces based on star charts, etc., culminating in one where he merely took the imperfections in the staff paper itself, and made them into notes?

Anyway I set out to explore the possibilities of appropriation in the realm of abstract digital audio.... could I build up a piece containing some impressive pre-existing gestures that I merely "discovered", where I had never actually "composed" any of it, never imagined, never wrote down, never programmed a single note of it..... but rather, appropriated it out of digital data that never began as "audio" in the first place? Took it all and shaped it, as a sculptor carves a new "work" out of a chunk of wood, making use of the existing structure in the grain of the wood? And could I make a real good piece out of it, one that would stand up to critical listening? And would it rock, would it shred? :-)

Note that I was NOT interested in "sampling" previous musical extracts, nor copying melodic material. I wanted to go down below to find musical gestures in raw data.

I found that there were software tools to change the specifications given in soundfile headers, strip them off entirely leaving only "raw" data, or even slap a newly created header onto what had previously been a "raw" file.

I set out to find datafiles in common places on various computers, which would provide interesting results when I would stick a soundfile header on them, interpreting their data as audio signals. I know you're all going to go and steal my idea now :-) but that's OK, since I "appropriated" the material in the first place... :-) (and since the piece is dated 1993 on my CD, preserved for all time, plus I did a lot of structural work on it, over and above the basic appropriation)

I found that a lot of files on various hard disks were uninteresting when interpreted as audio. The Macintosh finder, for example, just didn't do it for me, nor was it large enough to provide a long enough audio file to work with.

I found that the "swapfile" of the Mach operating system on a NeXT cube did have interesting possibilities though. NeXT's old non-microkernel Mach operating system pages to inodes, meaning that the virtual memory "swap space" is available as a Unix file (I guess Sun's Solaris operating system pages to inodes too; no other Unix systems could page to inodes in the old days, they needed to have a dedicated "raw" partition instead.

Due to a basic design problem in NeXT's ancient version of Mach, the swapfile grows and grows, but never shrinks, until either it fills the disk, or you reboot. So after a few weeks, it can get big and juicy, especially on a shared workstation seeing a lot of activity. (and most of the activity on this particular workstation was people editing audio soundfiles!! :-)

I collected the swapfiles for a while and put lots of different soundfile headers on them (stereo vs. mono, different sampling rates, 16-bit linear vs. floating point). Some awesome, frightening stuff was emerging. Gestures perhaps impossible, or at least extremely unlikely, with human performance were appearing, although they were mixed in with lots of uninteresting "static". Circuit breakers in studio JBL loudspeakers would reset :-)

I developed my little (actually rather large, in megabytes :-) collection of files which were now soundfiles. Then I fiddled around with them, using Paul Lansky's "RT" real-time mixing program on the NeXT. I incrementally evolved some very complex RT scripts, with all sorts of insane constant pitch shifting, panning, loudness changing, applied to a set of the soundfiles. So I got lots of new stuff, even more sublime than the raw power of the original soundfiles. Delicate, actually, some really wonderful subtle quiet nuances at moments.

And then I had all this wonderful STUFF laying around. People could wander by the studio while I was working on it, playing some back, and shudder at what was attempting to emerge from the speakers :-)

But now, now that I had my source material ready, what to do with it, how to "compose" with it, how to make a PIECE, a WORK, out of it?

So I began a process of selection, of auditioning and cutting/saving GOOD bits. Imagine, in the old days of analog tape, I would now have tables and drawers overflowing full of bits of tape, waiting to be spliced together bit by bit :-) (I recall John Cage's description of some splicing sessions, read on the "Indeterminacy" recording, about how he and his collaborator Earle Brown (?) found that they measured different lengths even using the same ruler.... :-)

A sort of "dramatic curve" began to take shape, in my mind, as I cut & pasted among my zillions of soundfiles into the main piece. The stuff from the crazy RT panning/hairpin loudness sessions comes first, some of the most beautiful subtle gestures appear; loudness & emotional intensity builds up, a valve opens which gives a taste of the cosmic throughput swoosh of the unadulterated original data, then it dies back down to near-nothingness.

At this point such ghostly quiet stuff leaks through that I literally frightened myself! there's a sort of recap, a reprise of the original material.

End of a sort of Part 1 of 2. Into the second part, more "straightahead", no more RT mixed material, just stuff direct from original soundfiles, although the presentation is sequentially in bits from a number of different soundfiles. Loudness contour gradually increases throughout.

In playing some of the soundfiles for our composition professor Alvin Curran at Mills College, he complained that it was too tame, he wanted more extremes, "Stuck notes!" etc.

OK, as a matter of fact.... there was a segment of one datafile where there was like a whole minute's worth of a steady tone, must have been constant values in the underlying data.... so that goes into the final mix, a tribute to Alvin. I like how it sort of gradually slides into this with several false tones first.

And then it's just a race to the finish line. I had made a preliminary decision to make the overall length around 5 or 6 minutes. I let the loudness contour get way up there. I had some real nice powerful violent bits available, almost like machine guns of varying timbres (note: recently I heard Richard Karpens nice piece using REAL machine guns and helicopters, commissioned for the 1995 ICMC).....

So the end is a real circuit-breaker-tripper. In the DAT version of the piece, these final peaks go up to the maximum amplitude, i.e. FFFF (whatever, you know what I mean). It's interesting to note that in usual DAT recording, if the values go up to the maximum amplitude, it generally means that the incoming analog audio signal was too high, and you get the awful "digital clipping" effect. But in my case, none of this material had EVER existed in the analog domain, so there was no "clipping".... merely maximum amplitudes, found in the data, as defined by the format specification in the prepended soundfile header (e.g. 16-bit linear or floating point)!

When I had the material mastered for the CD, I paid a qualified, experienced mastering engineer (Jason Rau of Monsterdisk in Chicago), and LISTENED to what he said (if you put out your own CD, you should pay a mastering engineer too).

He told me I'd be real stupid to leave those max amplitude sections on the CD. I was kind of sad but I agreed and he balanced the overall max levels on this piece with the max levels on other pieces of the CD. So it may not have quite the total overwhelming impact, in the CD version, as it did in the original concert premiere from DAT (for my MFA Thesis Concert, I got to use a wonderful pair of Meyer 733's (?) borrowed from Ralph the VP of Meyer by Maryanne Amacher, and they could handle anything, sweetly).

What the hell. I don't want to hurt your ears, anyway, I just want to give you a weird exciting feeling in them :-)

OK, I'll stop now :-) hope this isn't too long to get through the gateways onto your news spool...

Chris Koenigsberg ckk@uchicago.edu