Dan Zimmerman:

>Another issue that I am know worrying about is related to the question of
>how to classify a composition.

Several years ago I devoted far too much time to trying to solve
this problem, and not just for Western classical music.  I divided
all music into a number of basic categories, classified it on a
multi-dimensional system with axes for things like number of performers,
geographical/cultural origin, historical period, a "light to heavy" scale,
and others I now can't remember, and devised a mathematical coding system
that expressed degrees of hybridisation between one kind of music and
another.  I ended up with, IIRC, a 12-character code that could pin down
just about any piece of music from any period of any culture (except the
ones I don't know anything about, of course, but I included an updateable
"ignorance" factor to allow for them) and which also worked fairly well in
the reverse direction - give me a code, and I could describe what the music
would be like with a moderate degree of success.  And also distinguish a
"fake" code from a real one, most of the time.

My original motive for this exercise was to find a better way of
classifying my recorded music collection, which was much larger then than
it is now (I regret to say that it was a fair way on the other side of the
magic 2000 figure).  Since then my attitude to recorded music has changed,
and I now place a much higher priority on live music than on recordings,
which I think are rather like stuffed animals in natural history museums -
useful for study and as a reminder of the real living things, but very very
different from them in most of the ways that matter.

And so I rather lost interest in my grand scheme for classification of
all music.  I've probably got the details up in the attic somewhere, but I
don't think finding it is a good enough reason to go up there and disturb
all the spiders.

Ian Crisp
[log in to unmask]