If you need to know about computers and semiotics, Peter Bøgh Andersen is the fellow to see. I believe he works in the same department, now, as Peter Nürnberg, which must make for some interesting discussions. (Nürnberg is slated to give a much-anticipated keynote at Hypertext 03)
Andersen uses The Lord Of The Rings as an example narrative and makes a better job of it than Celia Pearce another student of games who relies on Tolkien. It's an interesting paper, suggesting just how difficult interactive narrative can be and laying out some interesting scaffolding for bridging the roughest sections of the narrative path. (Thanks, Klastrup!)
But there are problems. Andersen is (justifiably) worried about ensuring that interactivity is honest, that reader choices matter, and seizes on chaos theory as a possible salvation.
"If neither the game designer nor the consumer, for principled reasons, can predict the reaction of the system to their manipulation, then the reaction cannot be said to be intentionally caused by them."
This is intriguing, but Andersen proceeds to construct an illustrative example of chaos in a game setting that is, regrettably, not chaotic. The self-organizing narrative of Middle Earth is not, as far as I can see, self organizing. Why scare people with math and formal machinery, if you aren't going to actually turn the crank?
I wish I knew whether the absence of Genette (or Walker '99) from Andersen's discussion was principled, a sign of disagreement, or simply an assumption that of course we all know this. It's a short paper, perhaps confined by constraints of space or schedule, but I wonder about these and other missing voices. Someone has got to reconcile Michael Joyce's "Nonce Upon Some Times: Rereading Hypertext Fiction" (MFS 43 579-597) with Schankian formal (or formulaic) analysis: I think Andersen might be perfect for the job, and this might be the right place to do it.
Andersen does get most of his Tolkien right, or close enough that the summaries aren't unbearable. Some details are rough. Orcs are said to always act to oppose Frodo's quest; what about Shagrat? Indeed, Sauron's allies are always frustrating him and each other; "I don't trust all of my boys, and none of yours" is the language of Mordor. Andersen says that "few other objects can be 'claimed' in the book" as Sauron would claim the One: what of Anduril, Glamdring, Shadowfax, Sting, the Three, Durin's Bane, Khazad-dûm, the Throne of Gondor, or the Sackville-Baggins spoons?
More seriously, Andersen concentrates exclusively on Frodo's story. His analysis is silent, for example, on the linguistic depth of the trilogy. The class tensions that divide and ultimately unite the four hobbits -- mirrored later in the historical conflict among the Wild Men, the Riders of Rohan, and the descendants of Nûmenor -- don't fit into Andersen's schematics of motivations and agencies, any more than they fit into Schank or Propp. Perhaps this is asking too much of a new theory, but it might be a sign that the edifice will collapse at the first big storm. And if you don't understand why four hobbits set out from the Shire, you don't understand the story.
More seriously, in my opinion, Andersen skips entirely over the twin problems I call My Friend, Hamlet. First, the imaginative reader is bound to think of things the creator never envisioned, and the reader's best thinking inevitably generates the dullest response: "I don't understand." And, second, if we let a sane and sensible person like you onto the stage, everything may melt.
Tragedy requires that the characters be blind (as we ourselves, at times, are blind), and if you let a sane and sensible reader into the room, everything is bound to collapse. Take Hamlet: it's absolutely obvious that he should go back to school, get roaring drunk, get laid, and await his opportunity. He knows this. Horatio knows this, Ophelia knows this. Even Claudius and Gertrude know -- why else send for his college pals? Nobody can bring themselves to say the words -- that's the tragedy. But what's to stop the reader? Only brute force and error messages ("You can't do that") that call attention to the arbitrary boundaries of the world. If you make Hamlet a game, it has to be rigged.
Pearce thinks Tolkien is a story for kids, and I'm not certain that Andersen doesn't make the same misreading. On the one hand, he's clearly influenced by Schank's early work on representing childrens' stories, and he does mention reading Tolkien to his kids. On the other hand, though, Andersen knows and respects his Viking lore, and his closing speculations on Ní∂ingr in the user interface may be the most interesting aspects of the paper.