April 17th, 2010

My reply to Roger Ebert’s “Video Games can never be art“:

I want to write in response to your article because I used to work in the gaming industry, and my original interest in it was precisely because I wanted to make art. I left the industry mostly because I felt it had moved in a direction that was more or less antithetical to that desire; nearly every game being made today suffers from fundamental problems that prevent them from really rising to the level of art, even if they can be quite impressive games in their own way.

I agree in general with your dismissal of most existing games as art (at least as “good” art), I disagree, however, with your contention that games either cannot be art or won’t be art within the lifetime of living gamers. What I believe is that it is very difficult to make games which are art, but to assert that they cannot be art I believe is based on a lack of understanding of what differentiates a game from other existing art forms.

We can start with a reductio ad absurdum argument: A film could be thought of as a game without any choices. But something seems to go wrong the moment we add in choice: i.e., there does seem to be something fundamentally less satisfying, artistically, about a choose your own adventure novel than a regular novel. So obviously there’s something about interactivity itself which makes art problematic.

But not all interactivity. Suppose a great film critic (ahem) created an interactive library of films, complete with the option of selecting commentary or not commentary, rewatching certain film segments with and without commentary, watching films in an order selected by the user, and so on. One could argue whether or not curation is art but certainly I think it is plausible to suggest that collages, mashups, mixes can be considered art — and curating an interactive film collection I think could certainly rise to the level of art, provided the curation was clever enough. So I don’t think the problem is necessarily that interactivity itself dooms art.

But of course in this case, the individual elements of the interactive experience were not originally intended to be interactive. They each created an artwork that was intended to stand on its own. There is something about creating something noninteractive which makes it much easier to create art, and the interactivity in my example is comparatively unobtrusive and doesn’t affect the content of each piece. The choices don’t interact significantly with the worlds of each element (film); there is a natural relationship between the choices and what is being presented.

So this leads me to my theory about the problematic nature of interactivity and art. The problem, it seems to me, is that the way in which interactivity interacts with the game world creates a fundamental breaking of suspension of disbelief, because the interactivity itself forces a highly artificial and noticeable break from our expectations extrapolated from the nature of the world being depicted in the game. Part of the problem is that many games try to live within some sort of story world, or involve interactions with what appear to be creatures or characters. But creatures or characters in the real world behave in ways far more nuanced and complex than characters do in game worlds; the only thing we are able to simulate effectively and accurately is physics of objects, and thus games are getting more and more realistic in terms of their physics, but the interactions with characters remain completely unbelievable.

And this is not merely about the quality of writing. It has to do with the expectations of interactivity itself. When you are free to make choices in the game world that seem relatively unconstrained, but then you come across or interact with characters who allow you either no choices or just a few “branches” of choices, that is fundamentally at odds with our expectation. In the real world you don’t switch back and forth between modes where you can freely do things in the world and modes where you can only listen to canned speeches. Even if the speeches were brilliantly written, the jarring transition between interactivity and noninteractivity prevents the character interactions from being emotionally engaging.

The same, I believe, applies to games that don’t involve storylines, per se; even platform games that have animals or other creatures in them — because these creatures behave in ways that are extremely routinized, so there’s no real sense that you’re actually interacting with people or animals or real monsters. They’re just cutout figures for the most part.

I would submit, however, that earlier in the history of computer games there were games that I would say could have begun to qualify as a sort of primitive art — precisely because they were so limited. For example, consider Marble Madness — a game which was simple by modern standards and involved rolling a marble down fanciful surfaces. On the surface this game seems quite like every other similar game, but I’d say it could start to be considered a kind of simple, cartoonish, abstract art in its own way, precisely because the world was so constrained that one could easily totally immerse oneself in the world without a jarring sense of breaking of suspension of disbelief. The creatures you interacted with were abstract (a bowling ball, moving green ooze, etc.) and thus their movements and behavior were consistent with the artificiality of the algorithms that controlled them. The interactivity was constrained in a way that didn’t violate the implied rules of the world. In other words, it wasn’t an attempt to imitate life; it was an attempt to create a self-contained interactive world with its own properties that was self-consistent.

Other games which I think could begin to qualify would include things like Sim City — again, not really realistic, but in an abstract sense I think the game itself could be thought to be a simple form of art. The rules of the world are constructed in such a way that you are interacting with it at a level of abstraction where the artificiality of the constraints don’t slap you in the face. I wouldn’t say Sim City is a particularly great artwork (in fact I’d say Marble Madness is better as an artwork) but it is the sort of design which I think, executed very well, could become art in some sense.

The key, it seems to me, is to make sure that the interactivity is constrained in a way that, within the properties of the world you’ve created, within its limitations, feels natural (this does not have to mean realistic). Once you’ve done that, then I think you have at least the possibility of starting to create art.

What if you want to create a game that involves characters, storytelling of some kind? This is a much more difficult problem. Years ago I worked on a game design which was intended to address this problem. In this game, you could interact with characters — not by very simplistic or totally constrained conversations, but via a series of buttons. The buttons had labels like “Snub” “Flirt” “Kiss Up” “Insult” and so on. These buttons would change as the conversation proceeded. When you hit a button the game would show what you said, and then show what the artificial character replied. The interesting thing about the design is that people who played our prototype found themselves identifying with what their character said even though they hadn’t actually typed it out or said it themselves. They’d often find themselves shocked by the specific nature of the replies they received… as though the computer had “understood” them.

This differed from the common practice in some games of listing out several whole sentences and asking the player to choose between them. Doing that, you totally destroy suspension of disbelief; people don’t choose between sentences when talking to others. By only showing brief one- or two-word descriptions, we latched onto a psychological tendency in people to think about what they’re going to say in terms of intention before actually saying the words. The conversations were relatively realistic, aside from that. Even with this simplification it was very difficult to make these conversations work, but we managed to do it with clever reuse of lines of dialogue in different contexts and a lot of coding work. With a few hundred lines of reusable dialogue you could make a conversation which seemed funny and realistic that lasted 5-15 lines. We created what I still think is the only reasonable approach to doing interactive conversations with artificial characters.

There were other factors in our game: the “point” of the game was primarily the experience. We didn’t have puzzles. Sometimes there were goals, but we designed it consciously so that even if you failed to get your goal you still had a satisfying experience.

Unfortunately the company that funded our project (initially Disney and then Trilobyte) went out of business before they published the game, and there’s still a fair amount of work to do before it could be released. I still have the source code and the rights to the game, however, and our prototypes which work… I do hope someday to have the time or money to finish this thing. Will it be art? I am not sure the game could be compared to a Rembrandt, but I do think it could have been compared to a sitcom on TV (it was essentially an interactive sitcom). It is possible to make games art.

permalink | comments(rss)

comment trackback