[Humanist] 26.95 aesthetic computing

Humanist Discussion Group willard.mccarty at mccarty.org.uk
Mon Jun 18 22:16:42 CEST 2012


                  Humanist Discussion Group, Vol. 26, No. 95.
            Department of Digital Humanities, King's College London
                       www.digitalhumanities.org/humanist
                Submit to: humanist at lists.digitalhumanities.org

  [1]   From:    Jascha Kessler <urim1 at verizon.net>                        (31)
        Subject: Re: [Humanist] 26.94 aesthetic computing

  [2]   From:    amsler at cs.utexas.edu                                      (70)
        Subject: Re: [Humanist] 26.94 aesthetic computing


--[1]------------------------------------------------------------------------
        Date: Sun, 17 Jun 2012 16:00:30 -0700
        From: Jascha Kessler <urim1 at verizon.net>
        Subject: Re: [Humanist] 26.94 aesthetic computing
        In-Reply-To: <20120617205840.5AD4A1581CA at woodward.joyent.us>


I am learning a lot from this thread, and it is per se interesting.
 Notwithstanding, I remain somewhat irritated by the use of the very term
"æsthetic" sans recognition that the history of the theory or philosophy of
the Æsthetic realm remains blurred, not acceptably given, and not
understood. I suppose I refer to the absolute relativity expressed by the
old phrase: *B**eauty is in the eye of the beholder.  *Call it what you
will, eye, perceptor, phenomenal mass, whatever.  As each human being is
unique [pace Kierkegaard], and a singularity, as it were, or so to say,
from conception on, including identical twins, who part at delivery, first
and second out into the world, all of that being's perceptions, the
Æsthetic, will differ absolutely.  Maths and formulas in science might fix
one form or structure in time, and even then momentarily, and one can speak
of an "elegant" formulation or solution, and design of maximum
elegance...but I would think, if philosophical statements are bruited, it
ought to be recognized from the first that the underlying structure of the
Universe as an Existent is composed, so far as we can surmise, of particles
that flash in and out of Being itself.  Quanta, or Quantum statements,
statistical in nature, I think, suggest that.  So to speak of the
"Æsthetic" in programming is I would suppose altogether beside any useful
point, except for judging values, themselves...alas, momentary in the eyes
of beholders.
Jascha Kessler

On Sun, Jun 17, 2012 at 1:58 PM, Humanist Discussion Group <
willard.mccarty at mccarty.org.uk> wrote:

> the beauty of the code itself.

-- 
Jascha Kessler
Professor of English & Modern Literature, UCLA
Telephone/Facsimile: 310.393.4648
www.jfkessler.com
www.xlibris.com



--[2]------------------------------------------------------------------------
        Date: Sun, 17 Jun 2012 21:24:39 -0500
        From: amsler at cs.utexas.edu
        Subject: Re: [Humanist] 26.94 aesthetic computing
        In-Reply-To: <20120617205840.5AD4A1581CA at woodward.joyent.us>

I am curious about this discussion of 'aesthetic computing' (though  
the title was quite confusing for me and at first I thought the intent  
was to discuss programs whose products were to be aesthetic works,  
such as art or musical compositions).

I should give some of my background. I majored in mathematics at the   
undergraduate level and only switched to computer science during my  
second year of graduate school, getting a Masters in Computer  
Science/Mathematics. I then went on to get my Ph.D. nominally in  
computer science, but actually as an interdisciplinary degree using  
coursework in CS, Information Science and Anthropoligical Linguistics.  
I consider myself a computational linguist.

As a programmer and as a mathematician I am familiar with the concept  
of elegance in code (programs) and mathematical theorems. In both  
fields one of the criteria for elegance is the succintness of the  
statement of the program or the proof. When computers began generating  
proofs of mathematical theorems, typically using enumeration of  
hundreds or thousands of specific cases that led to a proof; those  
proofs were considered exceptionally inelegant. Atrocities actually.  
However, for both math and CS there is a special circumstance lacking  
in most of the humanities; an absolute yes/no result. Do the  
mathematical statements PROVE the theorem; does the code compute the  
desired output.

In programming (and math), the shortness of the statement is still  
almost all that is necessary to gain the appelation, "elegant". If  
someone can write a shorter program that computes the same result; or  
create a shorter proof of a theorem; that is a more elegant proof.

Shortness in programs is also a factor of something you haven't  
mentioned; the programming language used. This gets one into the  
question of high level languages vs. low level languages. It also  
raises the question of efficiency of the programming language (and by  
that I mean something separate from the elegance/shortness of the  
program). You see, alas, a programming language that you would use is  
translated into instructions for the computer in machine language. So,  
high level languages, which could permit you to write shorter  
programs, could translate into more machine language steps than lower  
level languages. That's why there are optimizers, programs that  
attempt to optimize machine language programs to contain fewer steps.

This makes the issue of aesthetic computing more complex; and the  
question of efficiency somewhat outside the hands of the programmer  
once he has chosen a given programming language in which to work.

However, the simplest statement here is that the more elegant program  
(or theorem) is always the shorter one that does the same thing. But  
that doesn't necessarily mean that program is the most efficient. Its  
elegance may be dependent upon the 'power' of the programming language  
that eliminates the need to specify more steps because its commands  
translate into more machine code instructions.

One final note. In programming, there is a separate concept of  
'readability' of code. This is often at the core of complaints or  
praise from programmers of other programmer's work. If a program is  
well-written it is easier to maintain because it is clearer what the  
code does at every step. Modifying such a program is much easier for  
subsequent programmers. Writing a program to be exceptionally readable  
is a special consideration; not unlike writing clearly in a natural  
language.

One thing programmers can do when they seek to conceal their code's  
function from being copied is to eliminate the use of mnemonic  
variables and comments in the code. If you systematically replace  
variable names like "INPUT" and "RESULT" with names like "X1" and  
"X2", the code still works---but becomes cryptic to those trying to  
read it and understand what it does. I don't know what to say about  
the effect such a translation would have upon the aesthetics of the  
program.

And, also, remember, the middle ground between programming and  
theorems is algorithms. Algorithms can also be elegant in the way  
mathematical proofs are elegant. The implementation of an algorithm in  
a programming language would seem to be a secondary form of  
expression. One could presumably have an elegant algorithm which was  
implemented as a far less elegant program by an unskilled programmer  
who selected the wrong programming language or wrote the code in an  
inelegant way. Elegance isn't necessarily guaranteed when you translate.





More information about the Humanist mailing list