===       ===     =============        ====
             ===       ===           ==            ==     =
            == ==     ===           ==           ==      ==
           ==   ==== ===           ==           ==      ==
          ==     ==  ==           ==            =       =
         ==         ==           ==             ==    == 
        ==         ==           ==               ====
       M U S I C          T H E O R Y         O N L I N E
                     A Publication of the
                   Society for Music Theory
          Copyright (c) 1994 Society for Music Theory
| Volume 0, Number 6      January, 1994    ISSN:  1067-3040   |
  All queries to: mto-editor@husc.harvard.edu
AUTHOR:  Stephen Smoliar
TITLE:  Comment on John Roeder's article
KEYWORDS:  Roeder, semiotics
REFERENCE:  mto.93.0.5.roeder.art
Stephen Smoliar 
National University of Singapore
Institute of Systems Science
Heng Mui Keng Terrace
Kent Ridge, SINGAPORE 1027
[1] As a computer scientist I have to confess that it is 
often difficult for me to stifle a choke whenever I 
confront the word "semiotics."  While I have the greatest 
respect for the heritage of the discipline, I also feel it 
is important to recognize that its pioneers were restricted 
by the intellectual equivalents of stone axes.  When they 
worked with symbols and codes, the only tools they had were 
their own pencils and papers.  That they were able to 
achieve as much as they did is admirable, but it more than 
a little tragic to think that there are still those who are 
blind to the limitations of those tools.
[2] It did not take long for the computer pioneers to 
recognize that their machines could manipulate symbols just 
as easily as numbers.  (Turing was certainly well aware of 
this, although I am not sure I would credit him as being 
the first of have the insight.)  Even when John von Neumann 
wrote his first computer program (documented by Donald E. 
Knuth in Volume 2, Number 4 of *Computing Surveys*) he 
realized that it was easier to work with symbols than "raw" 
binary code.  The need for computer programs which would 
process such symbolic codes was a very early insight which 
was quickly followed by the recognition that any such 
program would have to be based on models of the syntax and 
semantics of those codes.
[3] One might argue that work on machine translation of 
*natural* languages got off to a false start due to 
*ignorance* of semiotics--particularly the insight that 
there was more to language than syntax and semantics.  
Nevertheless, we eventually dug ourselves out of that hole.  
What is more important today is that we now have very 
powerful systems at our disposal which allow us to 
manipulate symbol structures;  and I find it a bit 
difficult to wade through much of the arcane language of 
the semiotic theorists when it seems to me as if it could 
all be articulated so much easier in LISP.  (This is not to 
say that LISP solves all problems.  Rather, it often helps 
us to formulate clearer questions, even when answering 
those questions turns out to be very difficult.)
[4] Having gotten all that off my chest, I would like to 
provide my own take on what "Toward a Semiotic Evaluation 
of Music Analysis" is really all about.  From where I sit 
(one hand on my Macintosh keyboard and the other at the 
piano), the key problem of music analysis is one of 
*description*;  and the key problem of description is that 
it cannot be readily reduced to a simple exchange of 
symbolic codes.  Put in the bluntest and most obvious way, 
a description of a music experience is not the same thing 
as a description of an algorithm.  When I need to describe 
an algorithm to my computer, I know that I have to respect 
certain *a priori* conventions assumed by the machine as to 
how I use my symbols to communicate.  When I need to 
describe a music experience, I cannot always assume that 
those conventions exist, let alone identify what they are.
[5] The task of description, as it applies to music 
analysis, is founded on two agents:  the agent giving the 
description and the agent receiving it.  Those two agents 
are rarely "of the same mind," so to speak.  (They 
certainly do not have identical mental states, no matter 
what particular philosophy of mental state you happen to 
subscribe to.)  Therefore, description is not so much a 
matter of the describing agent passing a code to the 
receiving agent as it is a matter of the two agents 
mutually negotiating towards a point where they have some 
confidence that they are both talking about the same thing.
[6] Within a community of experts (such as those who haunt 
meetings of the Society for Music Theory, for example), 
negotiation can often be minimized.  Music theorists 
constitute what Donald Schoen calls a "community of 
practice," within which *some* *a priori* conventions of 
communication can be assumed.  However, since most music 
theorists are interested in talking about new things, it is 
often the case that novelty goes beyond the limits of those 
conventions;  so even within a well-defined community of 
practice, the need for negotiated communication is always 
[7] From this point of view, Roeder offers a key insight.  
However, it is more an insight about negotiation than about 
codes.  Mathematics, graphical representations, and 
narrative are not codes which mediate our analyses.  
Rather, they are our negotiating tools:  different 
resources upon which we can draw when, in undertaking the 
task of description, we have to confront the problem of 
whether or not the receiving agent is really "getting it."  
Furthermore, if that receiving agent is not giving us any 
feedback (which is what happens when we are writing a paper 
rather than engaging in face-to-face dialog), then we might 
do well to consider how we may effect a *synthesis* which 
brings all three of those tools to bear.
[8] I am currently reviewing David Lewin's new book, 
*Musical Form and Transformation* for *Computer Music 
Journal*;  and I am trying to argue that Lewin has 
succeeded in such a synthesis in the four analyses in this 
book.  I mention this because I think one of the most 
important moments in Lewin's *Music Perception* paper, 
"Music Theory, Phenomenology, and Modes of Perception," is 
when he introduces the concept of "music behavior."  The 
descriptive act of music analysis is yet another instance 
of that behavior.  As such, we should think about it in 
terms of its behavioral implications rather than trying to 
reduce it to an exchange of codes.
Copyright Statement
[1] Music Theory Online (MTO) as a whole is Copyright (c) 1994,
all rights reserved, by the Society for Music Theory, which is
the owner of the journal.  Copyrights for individual items 
published in (MTO) are held by their authors.  Items appearing in 
MTO may be saved and stored in electronic or paper form, and may be 
shared among individuals for purposes of scholarly research or 
discussion, but may *not* be republished in any form, electronic or 
print, without prior, written permission from the author(s), and 
advance notification of the editors of MTO.
[2] Any redistributed form of items published in MTO must
include the following information in a form appropriate to
the medium in which the items are to appear:
	This item appeared in Music Theory Online
	It was authored by [FULL NAME, EMAIL ADDRESS],
	with whose written permission it is reprinted 
[3] Libraries may archive issues of MTO in electronic or paper 
form for public access so long as each issue is stored in its 
entirety, and no access fee is charged.  Exceptions to these 
requirements must be approved in writing by the editors of MTO, 
who will act in accordance with the decisions of the Society for 
Music Theory.