From newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.ecf!utgpu!csd.unb.ca!morgan.ucs.mun.ca!nstn.ns.ca!news.cs.indiana.edu!spool.mu.edu!uunet!zaphod.mps.ohio-state.edu!mips!darwin.sura.net!jvnc.net!nuscc!jit!smoliar Thu Apr 30 15:22:02 EDT 1992
Article 5193 of comp.ai.philosophy:
Xref: newshub.ccs.yorku.ca comp.ai.philosophy:5193 rec.music.classical:10378
Path: newshub.ccs.yorku.ca!ists!helios.physics.utoronto.ca!news-server.ecf!utgpu!csd.unb.ca!morgan.ucs.mun.ca!nstn.ns.ca!news.cs.indiana.edu!spool.mu.edu!uunet!zaphod.mps.ohio-state.edu!mips!darwin.sura.net!jvnc.net!nuscc!jit!smoliar
>From: smoliar@jit.iss.nus.sg (stephen smoliar)
Newsgroups: comp.ai.philosophy,rec.music.classical
Subject: Re: Intelligence, awareness, and esthetics
Message-ID: <1992Apr22.122656.1303@nuscc.nus.sg>
Date: 22 Apr 92 12:26:56 GMT
Article-I.D.: nuscc.1992Apr22.122656.1303
References: <1992Apr21.155531.23910@ncsa.uiuc.edu> <1992Apr21.212425.9210@saifr00.cfsat.honeywell.com>
Sender: usenet@nuscc.nus.sg
Reply-To: smoliar@iss.nus.sg (stephen smoliar)
Organization: Institute of Systems Science, NUS, Singapore
Lines: 98

In article <1992Apr21.212425.9210@saifr00.cfsat.honeywell.com>
shanks@saifr00.cfsat.honeywell.com (Mark Shanks) writes:
>>Shannon claimed that the symphonies as played by the "greats" produced higher
>>numbers than the same symphonies conducted/performed by less skilled
>>musicians.
>>
>>The claim was that the _rate_ at which information is transferred along the
>>channel to the receiver is higher for works by the great artists (and by a
>>few
>>institutionalized schizophrenics!) than it is for, say, black velvet Elvi.
>>
>>The implication was that it is completely possible to objectify aesthetics,
>>and
>>so justify a claim that some work of art was "better" than another...
>
>If, then, the "rate of transfer" is a quantifier of "value", then a modem 
>tone should outdo a Bruckner symphony or a choral work by Arvo Part. This
>is certainly specious reasoning. And on what basis was the "greatness" of
>the conductor determined? Was Toscanini "greater" than Furtwangler? Is
>Matisse a "greater" artist than Brughel? Is haiku "not as great" as Eliot? 
>Too many subjectivities to base an "objective" reading on. 
>
I have always felt that information theory does not get interesting until you
get past the Shannon entropy metric and start getting into serious problems of
coding.  Certainly the one time I heard Shannon talk, most of the talk was
based on various tricks by which you could milk all sorts of effect out of
only a few bits of communications bandwidth.  (This is still a popular subject
when you consider the current work in video compression.)  What makes all these
tricks work is the assumption that the receiver does not need all the bits of
the message.  Put another way, if the receive can make some set of expectations
about what the sender is going to transmit, then the sender only has to
transmit the bits that matter;  and the receiver can fill in the rest.

What does this have to do with aesthetics?  My own claim is that it is rather
important.  We tend to approach our aesthetic experiences with expectations
(just like everything else in the world from ordering food in Roger Schank's
favorite restaurant to the literary etiquette we anticipate in these messages).
There is a good example of this in an early scene of Peter Shaffer's AMADEUS.
Mozart has just arrived at Joseph's court, and Salieri has composed a march to
welcome him.  After all the formalities are over and Salieri offers him the
manuscript, Mozart claims that he has already memorized it.  He sits down at
the keyboard, rattles off the first eight measures or so, and then says, "It's
all just the same from here on, isn't it?"  The point is that Mozart knew
enough about how that sort of music was being written that he only had to
pick up a few bits (literally as well as figuratively) from Salieri to have
enough information to reproduce the whole composition.  This also explains
that old saw about Mozart's music having "too many notes."  Any critic who
voiced that complaint was basically objecting to having to cope with more
bits than his listening mind could tolerate.  The bit rate was high because
Mozart was NOT conforming to the expectations of his time, making the music
"harder" to listen to.  (We do not appreciate it because we are bombarded with
so much Mozart from birth that we have a different set of expectations.  For us
most of the bits come performances which basically upset expectations etched
into our memory by listening to the same performance over and over again.)

Hopefully this much should clear up Mark's paradox.  When we listen to a modem
tone, we really do not have any expectations;  so we are really not hearing the
bits that are encoded in those subtle shifts of the tone.  All we hear is a
carrier with unpredictable blips, and our expectations are not good enough
to characterize those blips.  Thus, the information content is about as low
as that of listening to white noise (high in the entropy metric but basically
information-free for aesthetic listening).  On the other hand, anyone with ANY
exposure to nineteenth-century music is going to have expectations when
listening to a Bruckner symphony (although the expectations you actually
have will vary according to your expertise in music history).  In the case
of Part, the bit rate will again depend on your knowledge of Part or your
general knowledge of past music.

This is where I think Shannon got it wrong.  Shannon seems to be assuming that
anything you want to say about transmission of bits has to do strictly with the
signal, i.e. the performed music (where "performing" may just be playing a tape
of electronic music or even a recording of more conventional music).  I would
argue that there is no bit rate inherent in the signal.  All we can talk about
are the bits received by any listener.  In other words (perhaps a bit like that
tree falling in the woods with no one to hear), "aesthetic signals" exist only
in the perceiving mind of the listener.  That listener has his own expectations
through which he decides what the bits are and how they are to be interpreted;
that interpretation is then his "aesthetic experience."

I would argue that this approach works even for the ultimate REDUCTIO AD
ABSURDUM, John Cage's 4'33".  Those familiar with the work expect to
"appreciate" the fact that there is no such thing as silence;  and their
ears pick up bits from all the ambient noise which occurs during the
performance (or, perhaps, even the lack of that ambient noise).  Those
who do not know the work find themselves flooded with bits which they
keep interpreting as meaning that the performer is about to play something;
and ultimately they end up as confounded as those who complained that Mozart
wrote "too many notes."

Let me close my saying that I am cross-posting this to rec.music.classical.
This has been a favorite topic of discussion there in the past.  If we really
want to try to discuss questions of information in music, it would be nice to
include some folks whose knowledge of music rises above the amateur level.
-- 
Stephen W. Smoliar; Institute of Systems Science
National University of Singapore; Heng Mui Keng Terrace
Kent Ridge, SINGAPORE 0511
Internet:  smoliar@iss.nus.sg


