March 27, 2006

E-FEST 2: First I Saw the Surface

by Nick Montfort · , 4:51 pm

At E-FEST 2006 I understood the main connection between postmodern writing and hypertext. It took being on a panel that included – along with computer scientist Lutz Hamel – George Landow, Stuart Moulthrop, and our own Scott Rettberg. This purported “Game of Fiction” panel was actually a veritable hypertext brainwashing session! I probably should have figured out this connection when I first read Landow’s classic Hypertext, just out in its third edition, or from Scott’s repeated statements of what I now seem to recall as this very point. But I think it was the comment from Robert Coover after the presenters spoke that finally made the Super Hypertext Club click.

As I’ve come to see it: When your starting point as a writer is language, and what language does as a complex and connected surface of words, it makes sense to engage the computer by using a technology that can describe a new topology of this surface. An underlying model of a simulated world, of the sort IF provides, is not the first nice capability that you’d desire.

In Coover’s 1968 The Universal Baseball Association, Inc., J.Henry Waugh, Prop., the main character is a creature who writes in a way that is exactly opposite from the postmodern novelist’s engagement with language and where it leads. He begins by simulating events in a world, tossing dice and consulting charts, and then narrates these copiously. But it would be a great surprise if Waugh himself was generated by this novel’s author using dice-based storytelling schemes. Everything suggests to me that Waugh is an anti-Coover, and that his “existence” was determined along the way as he was written, not the other way around.

Now, this realization doesn’t lead me to abandon IF, but it does remind me that a pipelined process of world building followed by text generation does not model the idealized postmodern writing process, and it doesn’t model my own, for that matter. This isn’t an argument against, for instance, the abstraction of simulation and narration in IF, but it does mean that it’s important to not assume that everyone will want to build a particular “fundamental” part of an electronic literary work first.

9 Responses to “E-FEST 2: First I Saw the Surface”


  1. Grand Text Auto » E-FEST 1: The Literary and the Arts Says:

    [...] few short notes about this year’s E-FEST at Brown, which was a great event. (Update: Note 2, “First I Saw the Surface;” Note 3, “Virtua [...]

  2. David G. Durand Says:

    I was at this talk, and was struck by this discussion as well. At the time I thought Nick was pretty committed to the world-building aesthetic, but I see now that he is agreeing with my comments at the time.

    If you enter the world of genre, whether SF, Historical novels, or Detective fiction, the world building notion seems more applicable. However, my impression is that the assumption that building a world is the first step to writing a novel tends to be a fannish conceit rather than an effective strategy. Hal Clement was famous for his use of fairly rigorous scientific analysis based on current knowledge to create the outer space worlds where his stories and novels took place, but even so, the stories themselves seem to come from a different place.

    Writers of these sorts of fiction tend to warn that research can become addictive, and an obstacle to writing.

    The joy of fictional worlds comes in large part from their incompleteness — No matter how much you learn about such a place, there is always a part of it that lives in your imagination, but no where else. That can draw you back to the text, looking for clues; can lead to the commission of fan-fiction; or can provide a sense of loss for what never was. At some level though, what is not described is more satisfying that what is known directly.

    Even where it resembles a system, fiction is just a tissue of lies, and lies are always shallow. Even the ideal game system engine is going to leak terribly, (to use Sapir’s description of grammars), and that’s OK, as long as the output in practice is good enough to fool someone.

  3. mark Says:

    The joy of fictional worlds comes in large part from their incompleteness — No matter how much you learn about such a place, there is always a part of it that lives in your imagination, but no where else.

    I don’t disagree, but I think the worlds are more interesting when they are more fully fleshed out, and seem to have some sort of self-consistency. That’s what, to me, separates something like Tolkien’s worlds from the legions of lesser fantasy. There is plenty left unanswered in Tolkien’s writing, but there is enough there that there’s some plausibility.

    It’s not just genre writing that’s like that either—Umberto Eco describes his writing process as world-building followed by an essentially random walk (he says once you have a world, you can just start positing that things happen, and each thing suggests another one that should happen after it, and before long you have a novel). Of course, Eco’s writing style is unusually digressive, but I’m not sure that’s a necessary outcome of his method (I think it has more to do with his enjoying lengthy philosophical asides).

  4. andrew Says:

    Regarding world-building, I made a similar argument, I think, about fictional character mind-building.

  5. noah Says:

    Mark, can I ask for a more detailed reference to Eco? I had heard something like that about The Name of the Rose, but when I went to look at Eco’s afterword it said something rather different — that it began with the desire to poison a monk, rather than with world building.

  6. mark Says:

    noah: I’m thinking of the section “The Novel as Cosmological Event” from “Postscript to The Name of the Rose” (that section is on pages 512-515 of the 1994 Harcourt Brace paperback edition of the novel).

    Some excerpts:

    What I mean is that to tell a story you must first of all construct a world, furnished as much as possible, down to the slightest details. If I were to construct a river, I would need two banks; and if on the left bank I put a fisherman, and if I were to give this fisherman a wrathful character and a police record, then I could start writing, translating into words everything that would inevitably happen.

    As you see, as soon as one’s invented world has been furnished just a little, there is already the beginning of a story. … The problem is to construct the world: the words will practically come on their own. Rem tene, cerba sequentur: grasp the subject, and the words will follow. This, I believe, is the opposite of what happens with poetry, which is more a case of verba tene, res sequentur: grasp the words, and the subject will follow.

    The first year of work on my novel was devoted to the construction of the world. Long registers of all the books that could be found in a medieval library. Lists of names and personal data for many characters, a number of whom were then excluded from the story. In other words, I had to know who the rest of the monks were, those who do not appear in the book. It was not necessary for the reader to know them, but I had to know them. …I conducted long architectural investigations, studying photographs and floor plans in the encyclopedia of architecture, to establish the arrangement of the abbey, the distances, even the number of steps in a spiral staircase.

    It is necessary to create constraints, in order to invent freely. … This has nothing to do with realism (even if it also explains realism). A completely unreal world can be constructed … but that world … must exist according to structures defined at the outset.

  7. noah Says:

    Mark, thanks for the excerpts.

    I don’t have the book with me, but I think you’re quoting from precisely the same postscript that gave me the opposite impression. The only line I wrote down in my notes is this one: “I began writing in March of 1978, prodded by a seminal idea: I felt like poisoning a monk.”

    I got the sense that Eco’s world building was in order to create the world for that story idea, not in order to generate a story idea via a “random walk” in the world.

  8. mark Says:

    Yes, I see that sentence as well. He does seem to suggest that the idea of a monk being poisoned got him the idea to write the novel, but when he describes the actual process of writing it, it’s very much “world first, plot later”. The impression I get is that he had a few vague scenes he wanted to be in there, and some other aesthetic feelings (like the fact that he’s a medievalist), and designed a world around those ideas. But during most of the design, he was doing minutiae like researching abbey architecture, drawing up long lists of plausible characters, and generally world-building. He seems to be asserting that he didn’t start writing the plot in earnest until he had all this done, and that when he did start writing, the pre-existing world is what made it easy: He could pick something that was interesting, and the detailed world constrained what things could plausibly follow; so he picked another interesting thing of those, and so on. So perhaps more like “greedy search” than “random walk”, but still different than the more plot-centric style of authorship, in which characters and locations are invented as needed, rather than “defined at the outset”.

  9. Dirk Scheuring Says:

    If he needs “a fisherman with a wrathful character and a police record” to start writing, while he himself “feels like poisoning a monk”, chances are that Eco does it like most experienced writers and starts with character. Tolkien, AFAIK, started with the imaginary languages he was developing, and created a world for those languages to exist in, but he’s a rare case (and in a way, starting with the languages that groups of characters speak is starting out with character, too). The “world” can usually be seen as a (hopefully) plausible expression of the motivations of the characters, which hinge on the motivations of the author(s), which (often) hinge on processes like “smoking in the bathtub”.

Powered by WordPress