Interactive Fiction, A.I., XYZZY, Brass Lanterns, Small White Houses, and Procedural Narrative
So as you know, one of my obsessions is with the notion of procedural narrative, also known as the automatic story machine, or the Novel that Writes Itself. As it turns out, other people have obsessions that dovetail (to a greater or lesser degree) with this wholly unrealistic and extravagant project. One group of artists that I am particularly fond of is the interactive fiction (IF) community, which I am glad to say is still chugging along, producing innovative literature in an utterly obscure and unappreciated field.
Lo, in the dawn of the Computer Age, there were no games but Lunar Lander and Spacewar, running on machines like the DEC PDP 11 and the IBM 3600. Darkness lay over the Earth, and there were none but bored genius engineering students with lots of time and access to mainframe computers on their hands. And so arose the First Games, and the greatest of these were Space Trek, Hunt the Wumpus, Hammurabi, and Colossal Cave (and of course all the other deserving pre-commercial games that I’m sure I’ve failed to properly appreciate). And many were those who perpetuated the underground distribution of this electronic samizdat, through dot-matrix printouts of computer code.
In the fullness of time, Colossal Cave begat Zork. And Zork became a commercially available computer game just as the Apple ][ and the TRS-80 began entering homes. And Infocom was born, along with dozens of other text-based computer game companies of all sizes and shapes. Great experiments were begun, and great empires were built, as fortunes were made and lost on games that had not one single bit of in-game graphical material.
But this state of affairs could not possibly last. Home computers were changing and evolving in unexpected and unpredictable ways. By the mid-80s …
Wait. I could digress for a really long time about all this history stuff. Look up “interactive fiction” in Wikipedia, and you’ll be up to speed. I wasn’t really intending to give you a comprehensive primer on the world of text adventures. What I wanted to do was discuss an important and far-reaching blog post by one of the leading lights of the interactive fiction (IF) world. Specifically, I’m referring to Emily Short (one of the founding authors of the post-commercial IF literary scene), and her post of February 13, 2008 on her Interactive Storytelling blog. (link: http://emshort.wordpress.com/2008/02/13/emergent-puzzle-solutions/).
There are a number of questions raised in both the entry and the resulting discussion about the nature of interactive fiction, the fundamental incompatibility of compelling interactive narrative with open-world sandboxes that are tediously simulationist, etc. Ms. Short is justifiably dubious about whether narrative can simply emerge from throwing enough dependent complex processes into a computer and then shaking vigorously.
Suppose that one wrote a story in which the large cast of characters could or might not interact with each other in all sorts of unpredictable but detailed and highly realistic ways. So a little village hums along with a hundred different randomly motivated autonomous agents. One agent gets hungry and goes to buy a loaf of bread, but meanwhile the baker has raised his prices to compensate for a grain shortage. On the other side of the village, the wheelwright is struggling to fit a metal rim onto a half-finished carriage wheel. The sooner the carriage is fixed, the sooner that the grain merchants will be able to negotiate and stabilize the grain futures market.
More often than not, the “stories” that would emerge from such a device would be incomprehensible, emotionally unsatisfying mechanistic messes; it doesn’t really matter how clever the game designers are, or how amazingly detailed and self consistent the physics, events, and objects in their games are. What’s important is the resonant literary transition must be more than a jumble of random puzzles and meaningless actions. The clockwork aesthetic is nothing more than flash unless it serves a broader story arc.
But I think Ms. Short is unfairly pessimist about the possibility of an ultimately unpredictable but emotionally satisfying narrative that is generated on the fly. One of the techniques is to code a limited set of types of stories, and then filter out all the boring inconclusive interactions, leaving only those interactions that fit in the narrow range of possible stories.