I don't play computer games. In fact, I feel kind of alienated by the whole genre. It's not that I object to violence, puzzles, or role-playing, or that I'm insufficiently awed by imaginative graphical worlds. On the contrary, I enjoy a good search-and-destroy mission in an out-lying galaxy as much as the next person. It's just that I'm more ambitious: given the choice, I'd rather create the galaxy than hunt the creatures in it.

What alienates me most about the computer games I've seen is that I'm asked to find my place in someone else's world. As challenging, intriguing, or beautiful as these worlds may be, I inevitably end up wishing I could get inside and change things, manipulate the very structure of the game to suit my taste. What bothers me, in other words, is that the player's imagination is almost always subordinate to the game designer's. The very success of the design provokes me. The better, more imaginative the game, the more I want to escape the player's subordinate role and play on equal terms with the designer-by redesigning the game.

Some current computer games allow players to customize the content or to create alternative scenes. There are numerous user-created levels for Quake and Doom, as well as personalized maps for Warcraft II and Civilization II. This is a step in the direction I mean. But the idea of incorporating the player's imaginative input in the functionality of a game could be taken much further. I imagine a game in which, in addition to customizing pre-existing content, the player could create entirely new games.

A game that emphasized the player's imagination as much as the designer's might be very different from what we are used to. At the same time, however, a game that treated its players as designers, rather than as "players," might appeal to the very large market of people alienated by action games, but eager for creative activity enhanced by the computer.

Games And Instruments

I don't play computer games. But I do play a musical instrument, and this kind of playing provides, I think, a powerful example of the imaginative freedom I'd like to experience when playing on the computer.

If we compare the computer to a musical instrument, then a computer game is like a piece of music, the game designer is like a composer, and the player is like a performer. As with a musical instrument, the computer offers an enormous range of creative possibilities. Each computer game, then, like a piece of music, takes advantage of this creative range according to a particular style and genre. The more talented the designer (composer), the more powerful and satisfying the game (composition). For the player, both games and music can be highly demanding manual and intellectual challenges. Playing a game, just like playing a piece of music, requires practice and skill. In both cases, the dedicated player is rewarded with the intense pleasure of immersion in a created world.

But the analogy between a computer game and a piece of music begins to break down for the player when we think about the expressive freedom of interpretation and performance. A player of Quake "performs" the game by devising unique combinations of moves within the fixed context of what the game allows. A pianist may perform a piece of music in a similar manner, creating a unique interpretation while still respecting the correct notes of a composition. The piano player can go further, improvising variations on existing music. Perhaps we could compare this to user-created variations in Quake. But the pianist can go further still, varying existing music so much that he or she really composes a new piece or even invents a new style or genre. Currently, the game player has very little to compare with this aspect of the piano player's freedom. Only by leaving the game itself and working with an authoring application can the player begin to make the transition from performer to composer.

This current restriction on the player's freedom is, I believe, a significant opportunity for game designers. Each limitation on the player's activity is also a limitation of the game industry as a whole. By thinking of the user as a potential composer, rather than simply as a performer, game designers could vastly increase the kinds of computer games available and, consequently, the kinds of people who buy them.

Activity Games

A computer game that thought of its player as a composer, rather than as a "player," would have to blur the boundary between a game and an authoring application. Once again, we can use the piano as a provocative example of what this might be like.

As an "authoring application," the piano is astonishingly economical. It creates infinite user interaction with an extremely limited range of materials: 12 notes, repeated at octaves, plus dynamics (the loud-soft or forte-piano that gives the instrument its name). Twelve notes, high and low, loud and soft. That's it. All made possible through the action of a single mechanical system: key-hammer-string-damper.

A computer game that took the piano as its model would probably have the following features:

• a limited number of "notes" or game elements ("gamesels");
• a simple means of combining these elements;
• a series of controls to modify the "dynamics" or inflection of these elements;
• and perhaps a repertoire of examples that players could use as starting points for their own creative expression.

Several games on the market already follow this basic model. Felix the Cat's Cartoon Toolbox, issued by Big Top in 1995, is an activity game that lets players create their own cartoons. In this game, Felix and his friends each come with a library of short animated clips isolating specific movements. By combining these clips, players create continuous action. A large selection of props, sound effects, music clips, and special effects makes the setting and mood infinitely variable.

A better known and more commercially successful example of the same basic idea is Barbie Fashion Designer from Mattel. Barbie Fashion Designer, which was the most popular CD-ROM game in 1996-97, lets players design clothes for Barbie dolls. Like Felix the Cat, it is a toolbox, a limited authoring environment in which the player combines simple pieces to create complex results. By selecting from one of 7 general styles (party clothes, work clothes, etc.) and then choosing different cuts, patterns, and ornaments, players can create an almost unlimited variety of fashions. A virtual Barbie will then model the clothes on a 3-D runway. Special fabric is included that works with inkjet and laser printers, allowing girls to print out and put together outfits without any sewing.

Both Felix the Cat and Barbie Fashion Designer are restricted authoring applications, in which the cast members have been defined, but their relations to each other are left open. Like a piano, these games achieve an extraordinary degree of interactivity with a small range of compositional elements. What characterizes both the piano and this kind of game is responsiveness to the player's imagination and skill. Like a musical instrument, they emphasize the user's creative expression rather than the designer's. These games treat their players as composers, rather than as contestants.

But when the user is given the role of composer in a computer game, then the analogy with a musical instrument shifts significantly. The game, instead of being like a self-contained piece of music, becomes like an instrument itself. And the game designer, rather than resembling the composer of a piece of music, now resembles an instrument maker.

If we think of Barbie Fashion Designer and Felix the Cat's Cartoon Toolbox as instruments, rather than as games, then designing and, above all, playing them takes on new meaning. And if we follow the implications of this expanded idea of "play" in relation to computer games, a new genre of "instrument-games" becomes possible.

Instrument Games

Both Felix the Cat and Barbie Fashion Designer can be compared to musical instruments because they let the user take the role of composer in relation to the game. By arranging a limited set of game elements ("gamesels") in endlessly new ways, the player uses the game as a medium of creative expression. This approach to games could of course be extended to many popular toys and fictional characters.

Favorite childhood toys like Tinker Toys, Lego, and even a chemistry set are already varieties of instrument-games. This approach could be easily carried over to other, more narrative forms. Letting the user play Toy Story or this season's Godzilla by creating new stories with these characters offers an alternative to the more conventional, arcade-style game that now predictably accompanies action films. Indeed, almost every popular book, film, television show, or traditional computer game could give rise to similar, creative products in which the user is invited to rearrange the characters and settings rather than follow a pre-determined plot.

Pushing this basic idea further, the game designer who thought of him- or herself as an instrument maker could explore the possibility of media instruments whose product would be multimedia rather than music. Felix the Cat's Cartoon Toolbox, for example, is a media instrument whose "music" is cartoons. It is easy to imagine a similar music video game, in which players could compose music or create videos for their favorite songs from a library of sounds and images, plus effects and transitions. Like a cartoon toolbox, this would function as a limited authoring application, directed towards a specific genre. Music videos, soap operas, romance novels-any conventionalized form could be material for a media instrument that endlessly recombines modular "gamesels." And, as Barbie Fashion Designer demonstrates, the "music" produced by these instruments does not have to remain on the computer. If Barbie Fashion Designer is an instrument whose product is clothes for Barbie dolls, other instruments could create other objects, from paper dolls, houses, airplanes and spaceships to puzzles, jewelry, and sculpture.

The essential distinction between a instrument-game and a full-scale authoring application is the focus on a specific product or genre. Like a piano, these games are designed to produce a specific kind of output. The piano, in fact, is very restricted in this regard, since it physically limits the player to the 12 notes of the Western well-tempered scale. Here it is similar to Felix the Cat, since this game also limits the player to very specific compositional elements. However, different instruments have different degrees of limitation.

A violin is less restrictive than the piano because it has no fixed keyboard; the violin can play many more notes than a piano. Yet both piano and violin are more restrictive than a synthesizer, because they each have a distinctive sound, while the synthesizer can produce the sounds of most traditional instruments and many non-traditional ones, like sirens or wind effects. The synthesizer, more than an instrument, is a "sound processor." The synthesizer player has control over an enormous palette of sound sources, in addition to the infinite range of combinations.

If we take the synthesizer rather than the piano as our example, then the instrument-games we can imagine become far more complex and experimental. Synthesizer-games would expand the idea of media instruments to include multiple input and output options, focused on a specific product or genre, but capable of a wide range of transformations. A cartoon toolbox that acted like a synthesizer would include characters from several different sources or in several different styles, perhaps even accepting input from print or video, as well as have the basic tools to alter these game elements. A Fashion Designer synthesizer might accept designs from magazines, in addition to the ones provided, or might operate on a larger scale, letting users create patterns for their own clothing as well as for their dolls.

We could even imagine a fully scaleable synthesizer-game that, on the simplest level, might resemble something as modest as Pac-Man or Tetris, but which included a library of more complicated characters with increasing orders of functionality, plus the ability to sample new "gamesels" from external sources, and a basic suite of tools to alter and recombine these elements. As with an instrument, users could begin by learning the simple tasks of the basic game, then, as they gain skill, move through levels of performance, until arriving at the role of composer. At this level, creating new scenarios, new game objectives, perhaps even new games altogether would be part of playing the game. It would be a game, an instrument, a synthesizer, and a medium of creative expression.

In this way, there is a progressive shift of emphasis from the game designer's imagination to the user's. Game designers still get to indulge themselves, and players still get the adrenaline rush of experiencing a created world. But the game does not enforce this relationship through its structure. Instead, it becomes an environment in which the designer and the player meet on equal terms.

Allowing both designers and players to adopt the role of composer seems to make sense as a business strategy as well. By offering a medium for the creative expression of all of its players, the instrument-game invites long-term imaginative engagement, beyond the one season of its initial popularity. And by encouraging a repertoire of user-created compositions, it promotes a community of players who compete by extending the scope of the game.

The comparison between a musical instrument and a game is really just an attempt to expand the meaning of "play" in relation to computers. The industry is still in such an early stage that even slight changes in how we define a "game" can have profound implications. The game designer who redefines "play" to include new activities like composition creates an entirely new branch of the business-and perhaps even a new art form. That's the kind of computer game that I'd enjoy playing.

Glenn Kurtz is a writer and multimedia consultant. He teaches Digital Media at Stanford University. (gkurtz@digitopia.com)