Pondering Norman's Question About Programs As Art

In the most recent edition of ACM Inroads magazine, Victor Norman penned a column arguing that the creation of a computer program can be an artistic activity, and that a computer program can contain beauty*. The article ended with the open question, unaddressed:  

“Can a computer program itself be considered a piece of art?”

Curious, I did some research on definitions of “art” as a noun, across a variety of sources ranging from the mundane (dictionary.com and Wikipedia) to the erudite (Stanford Encyclopedia of Philosophy). Wherever I looked, I found general agreement that art (the noun) is typically understood to apply to visual media. There was also general agreement that to be considered a work of art, an object should have unusual or highly significant aesthetic qualities. Beyond these points, there is (according to the Stanford Encyclopedia of Philosophy) some disagreement based on era, culture and other factors.

How does a computer program fit in to these definitions if we stick to the two points of commonality: visual media and significant aesthetic qualities? Let's see. A computer program can be considered an object, albeit not one you can touch. Norman and others before him have explored the aesthetic qualities of code. A computer program does not fall into the category of traditional media, although in some cases it produces visual media when executed.

Execution and its attributes brings up an interesting point.  

When assessing the candidacy of a computer program for admission into the fold of “art”, does one evaluate the code itself, or the product of running the code? 

Most recent discussions of programming and beauty, including Norman’s article, study the code itself: how it was constructed. Why is the product of execution not considered as well? After all, 

the purpose of code is to be executed. Code has no reason for existence if it is not executed. 

Unless…one wants to claim that the lines of text are in and of themselves art. Art as a Noun.

All of which brings us back to those unusual or highly significant aesthetic qualities. Norman makes the parenthetical observation that “…we could find beauty in a computer program (a beauty perhaps only appreciable by skilled artisans of programming).” I’m not sure whether that caveat shoots his argument in the foot. In commonly accepted definitions of art, to what extent are the greater aesthetic qualities intended to be appreciable by a wide public audience?

Wikepedia notes that until the 17^th Century, the widely accepted definition of art was broader than today, encompassing not just what we now refer to as “fine art” but also including crafts and the sciences. If we revert to this expanded definition of art, a computer program (and the result of its execution) will qualify. I think.

            What do you think? Under what circumstances (if any) does a computer program qualify as a piece of art?

*ACM Inroads 2012 March Vol 3 No. 1 pp. 46-48