TITLE: "Mechanistic"
AUTHOR: Eugene Wallingford
DATE: October 31, 2005  7:19 PM
DESC: 
-----
BODY:
In my recent post on
<A HREF="http://www.cs.uni.edu/~wallingf/blog/archives/monthly/2005-10.html#e2005-10-24T19_36_27.htm">
   Gerry Sussman's talk at OOPSLA</A>,
I quoted Gerry Sussman quoting concert pianist
<A HREF="http://www.performancerecordings.com/">
   James Boyk</A>,
and then commented:

<BLOCKQUOTE><EM>
<STRONG>A work of art is a machine with an aesthetic
purpose.</STRONG>
<BR/><BR/>
(I am uncomfortable with the impression these quotes
give, that  artistic expression is mechanistic, though
I believe that artistic work depends deeply on craft
skills and unromantic practice.)
</EM></BLOCKQUOTE>

Thanks to the wonders of the web, James came across
my post and responded with to my parenthetical:

<BLOCKQUOTE><EM>
You may be amused to learn that fear of such comments
is the reason I never said this to anyone except my
wife, until I said it to Gerry!  Nevertheless, my
remark is true.  It's just that word "machine" that
rings dissonant bells for many people.
</EM></BLOCKQUOTE>

I <EM>was</EM> amused...  I mean, I am a computer
scientist and an old AI researcher.  The idea of a
program, a machine, being beautiful or even creating
beauty has been one of the key ideas running through
my entire professional life.   Yet even for me the
word "machine" conjured up a sense that devalued
art.  This was only my initial reaction to Sussman's
sentiment, though.  I also felt an almost immediate
need to mince my discomfort with a disclaimer about
the less romantic side of creation, in craft and
repetition.  I must be conflicted.

James then explained the intention underlying his use
of the mechanistic reference in way that struck close
to home for me:

<BLOCKQUOTE><EM>
I find the "machine" idea useful because it leads the
musician to look for, and expect to find, understandable
structures and processes in works of music.  This is
productive in itself, and at the same time, it highlights
the existence and importance of those elements of the 
music that are beyond this kind of understanding.
</EM></BLOCKQUOTE>

This is an excellent point, and it sheds light on other
domains of creation, including software development.
Knowing and applying programming patterns helps
programmers both to seek and recognize understandable
structures in large programs and to recognize the
presence and importance of the code that lies outside
of the patterns.  This is true even -- <EM>especially</EM>!?
-- for novice programmers, who are just beginning to
understand programs and their structure, and the process
of reading and writing them.  Much of the motivation for
work on the use of elementary patterns in instruction,
as we try to help learn to comprehend masses of code that
at first glance may seem but a jumble but which in fact
bear a lot of structure within them.  Recognizing code
that is and isn't part of recurring structure, and
understanding the role both play, is an essential skill
for the novice programmer to learn.

Folks like Gerry Sussman and Dick Gabriel do us all a
service by helping us to overcome our discomfort when
thinking of machines and beauty.  We can learn something
about science <EM>and</EM> about art.

Thanks to James for following up on my post with his
underlying insight!
-----