Inspired by the CHI '95 panel on creativity, Mark Gross posed the following dichotomy:

"THE COMPUTER IS TOO CONSTRAINING! it does not provide the freedom to create; it limits what I can do" or

"THERE AREN'T ENOUGH CONSTRAINTS!; the computer lets me do anything I want; there's nothing to work against, say as opposed to working with ink on rice paper, the medium is very unforgiving; you must execute the strokes with skill and craft; the constraints are very exacting and this enhances creativity."

Mark asks, "What are we to make of this dichotomy? Which side is right? What can we as media makers learn from this?"

As part of the panel, Frieder Nake referred to the computer as the purely semiotic medium of the postmodern situation. Later, I posed essentially Mark's question to Frieder in the following terms:

1. Art theory since Aristotle has focused on the imposition of form upon material. The material offers "resistance" (in CS terms, constraints) which the artist forms, masters, overcomes or reflects upon (depending on the historic era or your theory of art).

2. As purely semiotic, the computer medium has no resistive matter, no constraints imposed by the nature of the medium. It is universal at least in the sense of the universal Turing machine. For instance, as the composer Varese noted when computers were first invented, electronic music is free of all the restrictions of traditional musical instruments.

3. Yet, as programmers we know how restrictive any given software environment really is. For instance, since the 1960's I have been looking for an electronic music system where I could create sounds defined by arbitrary mathematical equations and I still have not found one or had the time to build one. Most music software re-imposes the constraints of traditional composition and none allow me to do what I want.

Frieder agreed that the relation of the universality of software to the constraints imposed on any software creator are complex. This relation is qualitatively different from what it was with material media; it is harder to grasp.

Here are some initial thoughts toward an aesthetics of software:

1. Let us extend our model of domain-oriented design environments as a sequence of successively higher level systems: machine instructions, assembler, languages, DODEs, a DODE customized to a particular creative person's interests and knowledge base, a creative artifact along with its specific software supports. Here we see that a final creative work depends not only upon the abstract potentialities of computation (which may be relatively universal and unconstrained -- at least for certain kinds of creative work), but on a concrete technology base (a chip design with an instruction set, controlled by an operating system, compilers, applications, environments, etc.). This technology base is extremely constraining. It's constraints are not a result of nature -- physical or chemical laws and evolutionary patterns -- but of countless human design decisions. These decisions encompass the history of technology, which is a central process in our society. Much of the concrete technology base has become second nature to us.

2. Second nature is a social construct. Our technology is vigorously pushed in directions corresponding to the economic priorities of modern society. Computers and their software are designed to promote productivity, a socially-specific form of efficiency measured by reductions in human work time. The ways in which computers could empower people to control information or processes or could empower them to reflect more deeply on issues related to what they are creating are suppressed in favor of speeding up work. A famous animator shown during the panel rejected the use of computers because it eliminated the time for reflection in drawing animations. But is this an inherent property of computers or just the use that they are put to in modern society? Shouldn't computers give animators tools to create images they could not otherwise envision (and therefore inspire their creativity)? Shouldn't computers free animators from tedious chores (e.g., with automated morphing) so they have more time, concentration and energy for creative reflection and imagination? And don't we particularly believe that software should support reflection with tools that augment human intelligence? But, at least as seen in the vast majority of software in the marketplace, computers just seem to push people to work faster and faster -- frantically, stressfully, almost mindlessly. Just as the tv in our society turns us into consumers, the computer turns us into efficient machines. Second nature in these parts is not conducive to creative activities.

3. I recently demoed at Apple headquarters a modest proposal for supporting creative curriculum development among high school teachers. The bottom-line response was: the marketplace does not support software that places any demand upon the user to think. Everyone agreed personally that the proposed software addressed a major social need in a reasonable way, but they insisted as decision makers that it was not appropriate for the economic market. If Apple refuses to pursue software that makes us smart, certainly IBM and Microsoft will not produce tools that promote creativity.

4. Throughout the history of software development there have been people advocating the design of software to enhance people's creativity. These advocates have been effectively marginalized by the marketplace, either as shareware producers who cannot make a living, as academics who also train students in skills needed by industry, as visionaries who propose ideas that may someday be adapted to the needs of industry, or as tokens who humanize the image of a ruthless economy.

5. There is, surprisingly, no recognized category of the software artist. Other technological media have been pioneered by artists: printing, photography, film, etc. The creative heroes of computer culture are the entrepreneurs: individuals who have come up with breakthrough ideas for practical systems that are successful in the market. But the inventor of a better mouse trap is not an artist. By software artist I do not mean a musician who programs electronic music or a visual artist who programs displays of colors. I assume that a software artist would create programs. That is, software would be the end product, the art work itself, not its means. Such a software artifact might demonstrate some potential of software that had not previously been thought about, even though it was not a practical, money-making idea. It might be a reflection upon the software medium, demonstrating in code something about the nature of software. (I can not think of a good example of this off hand -- if I could I might become a software artist, possibly the first one.) It might be a critique (or a postmodern deconstruction) of the way that software development is currently skewed by market forces.

6. Where would a software artist's works be displayed? Not at a traditional museum. Is there a journal of software art? I haven't seen any software art in Leonardo, the journal of art and technology that consistently treats technology as a means for forming some matter. Where is an audience? What would the products look like physically: textual code, displays of executing code on a monitor? Does the audience need to be able to read code and/or to understand technical issues? Surely a purely semiotic medium should be displayed in a virtual environment like the Web. The Agentsheets Remote Exploratorium comes immediately to mind as an appropriate venue. Coincidentally (?) there are Agentsheets titles that have to do with art (color or sound patterns) and others that have to do with software (e.g., petri nets).

7. Mark wants to know, "What can we as media makers learn from this?" I think our group has been trying to do what needs to be done. I think that our heroes (e.g., Rittel, Ehn, Winograd, Schoen) have been pointing the way for some time, emphasizing that we are designing media of communication and creation within a political (socio-economic) context. All we can do is to try to dream up and promote software approaches that empower people's personal creativity by supporting processes of reflection-in-action. Gerhard emphasized this during the panel. Of course, I would add that reflection or interpretation is from different perspectives and so we should support the tailoring of software to individual (not just domain) orientations.

8. Most software is designed to force the user to adapt to the software, so that the software serves as a management tool for controlling employees who use the software. In HCI we pretend that we are tuning software to the end user. But I think that if one analyzed the constraints on software design (even in American examples of so-called participatory design), one would find that corporate (Fortune 500) management interests overwhelm all other concerns. What can we do in the face of this? At least we can try to become more conscious of what we are doing and of the context in which we are doing it.

9. Software may be universal in a sense not true of any material medium. However, the second nature of actual software environments impose impenetrably complex constraints. There is an ideology that says technology is an impartial tool. There is even an ideology that says that marketplace decisions are necessarily the best and fairest. Unfortunately, we have no software artists who can poke holes in these ideologies. It is up to us to resist the prevailing ideologies and forces in order to push software design in directions that can allow users to be creators rather than just machines and consumers.

I assume that our efforts to address the needs of serious professionals such as designers is a useful antidote to treating users as consumers or robots. I assume that our interest in providing end-user languages, tailorable controls, evolving knowledge bases, media of interpersonal communication and tools for collaborative domain construction point in the right direction.

If we can come up with convincing examples of software that promotes personal creativity and unhurried reflection, then some people may demand such tools. Software artists, if they ever appear, may be able to demonstrate such potentials of our universal semiotic tool that have been largely suppressed to date but that we know could exist.

Gerry Stahl . . . from the philosopher's corner

Go to a response from Frieder Nake

Go to top of this page

Return to Gerry Stahl's Home Page

Send email to

This page last modified on January 05, 2004