I think the model I'm working with is a carnival. Unlike a movie, you
don't start at the beginning of a carnival and go to the end. Instead,
you encounter different "micro-narratives" within a carnival, at your
own pace and based upon what interests you. You, for example, go to the
tent with Magnifico the Mind Reader. When you do, you enter a narrative
with it's own structure, but even then, the narrative depends on
interaction with the audience, who can direct the course of the event.
Whenever it's over, you move on to the next event that interests you, and
engage a different narrative. Periodically, an event might engage you,
instead of vice versa, but even then, how you reacted would affect the
way it transpires.
I think this is a decent model for how an interactive work (whether VR
or multimedia) ought to operate, to separate it from linear media. I
haven't had a chance to play with Monty Python's Complete Waste of Time,
but I understand it works kind of like this (with more limited
interaction).
_Ought_ to work? Perish the thought! What you are describing is a
bunch of unrelated short stories under one roof. It's certainly the
most readily available model, and the easiest to implement. But it's
hardly the limit of what is possible. I'm not saying the approach is
invalid, just that it's exactly the sort of user interaction that I'm
suggesting we should try to move beyond.
Even a carnival has an implicit linear narrative. You enter, and in
no particular order, you go on all the rides. You pay a fee at each
one, and you have a pre-specified interaction with each one. Then,
when there are no more rides, you leave. The amount of interest value
generated, is linearly proportional to the number of short stories you
write up.
What I'm after, is something that exploits the _recombination_ of the
elements, so that a truly non-linear plot is developed
_in_the_viewer's_mind_. The viewer is the one who will create the
narrative, because the viewer will automatically attempt to relate and
narrative-ize what is seen in order to make sense of it. The
non-linear VR author's job is _not_ to provide narratives, but to
_direct_ the process of the viewer's free-form associations.
This is what I meant by attempting to understand the Surrealist
relations between objects. Steve suggested a theorem, that the
"semantic distance" between objects is related to their Surrealistic
impact. While it's not a complete explanation, it's a good start, and
a direction we need to pursue. We can't simply rely on random
collections of objects to do our work for us. I suspect that there is
some form of Surrealistic narrative principle underlying all of this,
whether it be the "dialectic of the Absurd," or the "paranoic-critical
system," or whatever. Once I've dug through a few more Surrealist
texts, I hope I can elaborate further on this.
Cheers,
Brandon