missed email

[Chris Crawford <chriscrawford@wave.net>]

Title: missed email

I had sent this out earlier, but to Morbus Iff rather than the entire group. I never do seem to get these things right. I hope that this addresses several of the concerns expressed since.
---------------------------------------------------------------------------------------------------------------

I'm not really happy with either of your examples, although they both do describe a single case of storytelling, with limited interaction. A better example of interactive storytelling is the guy setting out to tell a story with the kids interrupting, asking questions that require the story to develop in new directions, or perhaps expressing dissatisfaction with some aspect of the story. If the storyteller responds to these interruptions in a positive manner, then we get interactive storytelling.

ToastyKen asks: "But if the story changes due to this interaction, is not the story interactive?"

It's not the story that changes; have you ever seen a book rewrite itself? The story doesn't change, it is changed by the storyteller. If the storyteller changes the story due to this interaction, then the storytelling process is interactive. The idea I'm pushing is to focus on the process, not the data.

Finally, you're right in pointing out the repetitive nature of the "Tinkertoy Text" that I propose. It's pretty bad, but it's the best that I've seen actually implemented. As you say, with limited knowledge domains it might now be possible to apply some AI.

Chris