(Pierre Divenyi )

Subject: 
From:    Pierre Divenyi  <marva4!pdivenyi(at)UCDAVIS.EDU>
Date:    Wed, 15 Sep 1993 12:21:04 PDT

ucdavis!iss.nus.sg!smoliar ucdavis\!auditory(at)VM1.MCGILL.CA
Date: Wed, 15 Sep 93 10:41:39 PDT
From: marva4!pdivenyi (Pierre Divenyi)
Message-Id: <9309151741.AA24177(at)marva4.UUCP>
To: ucdavis!iss.nus.sg!smoliar ucdavis\!auditory(at)VM1.MCGILL.CA
Subject: Re:  Segmentation

While I am one of those having definite reservations as far as
the Lerdahl-Jackendorff system is concerned, we still have to
admit that the great bulk of music (in its broadest definition)
consists of discrete units (tones, attacks, etc.). The
Mussorgsky example of Steve Smoliar only raises the question
whether the smallest played units should be considered
the REAL units (in a musico-perceptuo-cognitive sense).
Thus, the problem to be solved is to establish units of
t i m i n g  -- for the sake of musical meaning and perceptual
organization. Grouping rules cannot (and should not) ever be
drawn up in disregard of such timing units. Yes, the
Mussorgsky notes in the last measures of Gnome are audible
only to the very best trained ears (I used to be capable of such tricks
thirty some years ago...) and, therefore, we can assume that the
timing units are quite a bit larger than the eight-note element.
Larger chunks would also work for music with uncertain unit
onsets, such as the Schoenberg piece cited, or the "composed
harmonic structures" (harmonics of a complex played by different
instruments) in pieces of composers like Gerard Grisey.

The 64,000-dollar question is: what are those larger timing units
and how to determine them experimentally. The hunch is yours!

        Pierre Divenyi