Bill Pirkle wrote:

>Having written software that composes classical music, ...

IMO one of the things which distinguishes CM from other musical traditions
is that each work has a composer.  Jazz is improvised; folk songs are
handed down from generation to generation (and sometimes improved in the
process); CM is written down, the presumption being that the composer
considered the finished composition to be good enough to be worth
preserving for posterity.

On this basis, computer software cannot compose CM.  Without judgement, the
software cannot decide what is worth preserving.

Possibly you meant to say that you had written tools for "computer-assisted
composition".  It's reasonable to expect a program to be able to solve
puzzle- canons, or even to suggest contrapuntal combinations of two or more
themes - that's the sort of thing computers do.  Also, IIRC there's a book
by Hindemith called something like "The Art of Tones" which ascribes levels
of tension to particular chords.  Using that, it shouldn't be hard to write
a program which could produce output corresponding to instructions such as
"give me ten minutes' worth of chords which start and end peacefully in C
major and have a very tense climax in C# minor two-thirds of the way
through".  The distinction is that there is still a (human) composer making
the judgement as to whether the chord progression is any good, and whether
a contrapuntal combination of the first and second subjects is appropriate.

Back to the question ...

>...  I am wondering if such music will ever be acceptable to CM types.
>I was wondering what about CM makes it appealing to people and have
>determined the following possibilities:
>
>Its historical value - that it was written in another time

Irrelevant - I'm not a historian.  BTW, I'm told (by a historian) that it's
a mistake to view historical people as living in the past.  Schubert didn't
think to himself "I'm living two hundred years ago, so I should compose
old-fashioned music for reactionaries".  Schubert composed music in the
present, and different people in different presents have enjoyed it ever
since.

>Its musical value - that is it musically well done, emotional, pleasing, etc.

This is everything, but tells you nothing.  I can, given time, come up
with several rationalisations about why I think Sibelius's music has value.
Admirers of Bartok could doubtless do the same about Bartok's music.  The
two sets of rationalisations may even have sentences in common, and that
might mislead the careless into thinking that we're saying something
profound and universal when in fact all we're saying is "I like Sibelius
and they like Bartok".

>Its composer - that it was written by, for example, Beethoven, Bach,...
>Its performance value - that it was performed by, for example, Rubenstein

These don't affect whether or not I like a piece.  However, they can make
a big difference as to whether or not I first listen to it.  When I hear
something that I like, I look out for other pieces by the same composer,
and sometimes by that composer's friends, pupils and teachers.  I like
Madetoja's 2nd Symphony because I like it, and looking for logical reasons
why I like it is missing the point.  I chose to investigate Madetoja's
music because he was a pupil of Sibelius, and it's possible that the
hypothetical Bartok-admirers have never heard any of it.  They're missing
out.  By the same token, it's likely that there are pieces by minor
Hungarian composers I'd enjoy if I were ever to hear them.

>These four, there are possibily more, should add up to 100% of its value.

Why should they? Irrational personal preferences don't have to obey the
rules of arithmetic.  Suppose that I enjoy a recording by a provincial
orchestra of a piece of no especial musical merit or historical
significance by an obscure composer.  How would this scheme cope?

Peter Varley
[log in to unmask]