Travel with me back to 1982. We are in a car. It is dark out, and we are parked in a wide open area with a giant silver screen before us. We are not alone, but not just because we are surrounded by other people in other cars. No, we are here to participate in a phenomenon. We are here to join the millions of people who saw E.T.: The Extra-Terrestrial.

Not only was this the highest-grossing movie of all time for well over a decade, it was also a wonder of licensed merchandizing. The cute-ugly alien showed his face on everything from lunchboxes to underwear. Slap E.T.'s squashed little mug on anything, and watch it sell like hotcakes. Well, maybe not anything.

One particularly infamous product to bear E.T.'s image was a videogame for the Atari console system. Sales figures were projected so high that more cartridges were manufactured than there were Atari units sold. Upon the game's release, audiences stayed away in droves. In the end, a warehouse full of neglected E.T. games was finally returned home, to a landfill, and then run over with a steamroller.

The moral of this story is that even with a hugely successful backstory that had essentially become part of American cultural heritage, this game failed. Of all the factors that lead to this downfall, I'm certain that being associated with E.T. was not one of them.

It should be news to no one who reads this that when videogames are analyzed these days, they tend to be analyzed as if they were film. When critics bemoan the state of the modern videogame industry, nigh-invariably do we hear the call for better stories. Better stories will make better games, we are told. But, at the risk of preaching to the choir, videogames are not cinema. Videogames are interactive. Interactivity makes them different, it makes them fun, and according to some it makes them dangerous. It is a terrible shame, therefore, that interactivity is not regularly examined for its artistic merits, despite having a provable aesthetic, albeit one that is currently only discussible in rather primitive terms of "good" and "bad."

MIT professor Janet Murray's Hamlet on the Holodeck attempts to provide a new Poetics for the discussion of "new media." Most relevant to the notion of interactivity is her concept of "agency." While Murray goes on at some length describing what agency is, I suggest the working definition, agency is the degree to which you are allowed to interact in a meaningful way with something. Although I do not wish to abandon my more general term of "interactivity," I believe that Murray's point of finding "meaningful" activity is key to making aesthetic judgments of the interactivity of a game. The more I am able to perform meaningful interactions within the scope of the game, the better the game's interactivity.

Good interactivity does not require that I be able to do literally anything at any time. Rather, my ability to interact with the environment needs only to match my desire to interact with it. A good level designer provides clues as to how I should interact with the environment. If the clues are subtle and well done enough, I shouldn't even realize that I am effectively being lead around by the nose.

When we notice a particular game isn't letting us do whatever we want, we say the game is putting us "on a rail." Reflexively, we seem to long for a game that grants us the ability to do anything we want. However, even the briefest study of the art of interactivity reveals that this is not what we want at all. We don't want to do everything; we want to do the right thing, or perhaps occasionally the cool thing.

Take for example actual pencil-and-paper roleplaying games. Here, the game's vocabulary theoretically as large as the combined vocabularies of all the players and the gamemaster. Yet even when we actually do anything we want, with magical powers or super technology to bypass even the steadiest physical laws of our reality, we still want a limited subset of those things: the right thing, which will advance the story, and the cool thing, which will strengthen the image of the character who portrays us.

When we play a videogame, we are limiting ourselves. Like a parent teaching a young child, the game limits our options to a set of condoned choices until we learn what the expected responses should be. The environment of the game is like a set of nouns, while the controls are like verbs. The better our set of verbs fits our set of nouns, the better a game's interactivity.

Let's look at a game where this idea is not just an analogy: the classic text adventure Zork. In Zork, I am an adventurer exploring a dungeon. I can interact with the environment by typing in commands describing what I want to do. There literally is a set of verbs and a set of nouns understood by the game. "Kill troll" works, while "Disembowel troll" does not. The problem felt by Zork, and many of its fellow text-adventures is that the game's vocabulary is not symbolic, it is literal. There is no punch button, there is a punch command. Although there may be no difference between a plank and a board, due to limited text-parsing ability, one may be a recognized noun, while the other is meaningless gobbledygook.

Another game with a much simpler vocabulary is Asteroids. There are basically three verbs: turn, thrust, shoot, and two nouns: asteroid and ship. Although my vocabulary in Asteroids is very limited, the game's interactivity works very well, because there isn't much else I want to do when I play Asteroids besides turn my ship and shoot asteroids.
The modern game Grand Theft Auto: Vice City, despite its reputation for being an open sandbox of a game where you can do whatever you want, has a surprisingly limited vocabulary. I can walk, run, crouch, jump, punch, shoot, drive, and hijack a vehicle. This is a good set of verbs to tell a gritty crime story, but if I decided that the plot should be about a thug who decides to give up his life of crime to start a bakery, I'd be hard pressed. Jim Munroe pokes fun at the disparity between GTA's limited verbs and its "do whatever you want" reputation in his short film My Trip to Liberty City, where he tries to force the thuggish main character into the role of a Canadian tourist who walks around Liberty City admiring its parks, and occasionally punching people "by accident." GTA does have a few options for the reformed criminal, however. If I can get into a taxi, I can make money by picking people up and driving them to their destinations. Of course, in order to get a taxi, I probably have to steal it, but perhaps this is simply taking the lesser of two evils.
The act of assembling our vocabulary of nouns and verbs is where story and interactivity meet. The story of a game is communicated through this set of words, with the slack taken up by cut-scenes (in a perfect marriage of story and interactivity cut-scenes would be unnecessary).

Without a doubt, the most common verb in gaming vocabularies is "go." Navigation of virtual space occurs in just about every game you can play, from Pong to The Secret of Monkey Island to Battlefield 1942. Moving around can be a great source, or the only source of fun in a videogame (Pong really only has one verb, "go"), but it can also be a real chore, as anyone who has ever spent several minutes traveling over non-descript plains and forests between towns in an RPG can attest to. Like any interaction, movement must be meaningful to be a "good" interaction.

About the worst use of movement in a game is, unfortunately, one of the most attractive to neophyte game designers, the maze. A person in a maze has two states: being lost, and being done. There's an old saying, "We're not lost, we're on an adventure!" If you put somebody into a maze, 99 times out of 100, they'll feel lost, and if they're lost, they're not having an adventure.

The great Argentinean writer Jorge Luis Borges wrote several stories about mazes. One of them featured two kings, one Arabian and one Babylonian, and their mazes. The Arabian king was visiting the Babylonian king, who had a great and complex hedge maze. He tricked the Arabian king into entering the maze and getting horribly lost, much to the embarrassment of the Arabian king. Some time later, the Babylonian king was visiting the Arabian king, who offered to show him his maze, which he said was even more difficult. He lead the Babylonian king out into the middle of the desert and left him there with nothing but sand on the horizon. Borges understood that a maze is not defined by twisty passages, all alike; it is defined by its ability to lose people. A maze may as well be a barren desert.

So how do games like Pac Man, NetHack, and Grand Theft Auto manage to be so successful when they are all about basically navigating twisty passages, all alike? They fill the twisty passages with interesting things! Mazes generally have only one point of interest, the exit. Filling a labyrinthine structure with interesting nouns, and equipping a player with appropriately interesting verbs not only it a more interesting place to be, it also creates instant landmarks.

Good interactivity isn't something that we've been seeking for very long. This is, perhaps, because it was only until we had finally invented rigidly non-interactive entertainment, that the idea of interactivity became important. It has been said that we use the term "interactive" only to refer to things with extremely limited interactivity, as opposed to none at all. After all, a dog is more interactive than the most advanced computer game you can name, but to call a dog "highly interactive" seems silly.

"Interactive" is currently something of an aging buzz-word. Since the rise of videogames, interactivity has been attempted across all sorts of entertainment: movies, television, video cassettes, books, theatre, to name a few. By and large, interactivity did not lead to any massive revolutions in these established media types, but the hype surrounding the promise of interactivity lead people to try combining with anything they could. We no longer see many experiments in interactivity in these mediums, as the hype has been replaced by disillusionment, but videogames, which by their very nature are interactive, continue to thrive, and techniques are beginning to develop within the field to user interactivity for more and more novel purposes. It is probably through videogames alone that "interactivity" can be truly studied.

If what I have essayed here has merit, if interactivity is coming into its own as an art form worthy of analysis, then I believe that one of the tasks that lies before us now is beginning a taxonomy of interactivity, a catalog of qualities by which interactivity can be qualitatively judged and analyzed. If we can do this thing, we can change the way videogames are talked about, and if we can change that, we will change the way they are seen, and ultimately the way they are made. This is my call to action. I look forward to seeing what comes.

Borges, J., The Complete Short Stories of Jorge Luis Borges
Munroe, J., My Trip to Liberty City [http://www.nomediakings.org/mytrip.htm]
Murray, J., Hamlet On the Holodeck: The Future of Narrative in Cyberspace
Sakey, M., There Are No Words (Yet): The Desperately Incomplete Language of Gaming [http://www.igda.org/articles/msakey_language.php]
Tmenezes, Are Games Art? [http://www.kuro5hin.org/story/2002/9/10/72851/0039]